Guide to Communication Research Methodologies: Quantitative, Qualitative and Rhetorical Research

research methods for communication

Overview of Communication

Communication research methods, quantitative research, qualitative research, rhetorical research, mixed methodology.

Students interested in earning a graduate degree in communication should have at least some interest in understanding communication theories and/or conducting communication research. As students advance from undergraduate to graduate programs, an interesting change takes place — the student is no longer just a repository for knowledge. Rather, the student is expected to learn while also creating knowledge. This new knowledge is largely generated through the development and completion of research in communication studies. Before exploring the different methodologies used to conduct communication research, it is important to have a foundational understanding of the field of communication.

Defining communication is much harder than it sounds. Indeed, scholars have argued about the topic for years, typically differing on the following topics:

  • Breadth : How many behaviors and actions should or should not be considered communication.
  • Intentionality : Whether the definition includes an intention to communicate.
  • Success : Whether someone was able to effectively communicate a message, or merely attempted to without it being received or understood.

However, most definitions discuss five main components, which include: sender, receiver, context/environment, medium, and message. Broadly speaking, communication research examines these components, asking questions about each of them and seeking to answer those questions.

As students seek to answer their own questions, they follow an approach similar to most other researchers. This approach proceeds in five steps: conceptualize, plan and design, implement a methodology, analyze and interpret, reconceptualize.

  • Conceptualize : In the conceptualization process, students develop their area of interest and determine if their specific questions and hypotheses are worth investigating. If the research has already been completed, or there is no practical reason to research the topic, students may need to find a different research topic.
  • Plan and Design : During planning and design students will select their methods of evaluation and decide how they plan to define their variables in a measurable way.
  • Implement a Methodology : When implementing a methodology, students collect the data and information they require. They may, for example, have decided to conduct a survey study. This is the step when they would use their survey to collect data. If students chose to conduct a rhetorical criticism, this is when they would analyze their text.
  • Analyze and Interpret : As students analyze and interpret their data or evidence, they transform the raw findings into meaningful insights. If they chose to conduct interviews, this would be the point in the process where they would evaluate the results of the interviews to find meaning as it relates to the communication phenomena of interest.
  • Reconceptualize : During reconceptualization, students ask how their findings speak to a larger body of research — studies related to theirs that have already been completed and research they should execute in the future to continue answering new questions.

This final step is crucial, and speaks to an important tenet of communication research: All research contributes to a better overall understanding of communication and moves the field forward by enabling the development of new theories.

In the field of communication, there are three main research methodologies: quantitative, qualitative, and rhetorical. As communication students progress in their careers, they will likely find themselves using one of these far more often than the others.

Quantitative research seeks to establish knowledge through the use of numbers and measurement. Within the overarching area of quantitative research, there are a variety of different methodologies. The most commonly used methodologies are experiments, surveys, content analysis, and meta-analysis. To better understand these research methods, you can explore the following examples:

Experiments : Experiments are an empirical form of research that enable the researcher to study communication in a controlled environment. For example, a researcher might know that there are typical responses people use when they are interrupted during a conversation. However, it might be unknown as to how frequency of interruption provokes those different responses (e.g., do communicators use different responses when interrupted once every 10 minutes versus once per minute?). An experiment would allow a researcher to create these two environments to test a hypothesis or answer a specific research question. As you can imagine, it would be very time consuming — and probably impossible — to view this and measure it in the real world. For that reason, an experiment would be perfect for this research inquiry.

Surveys : Surveys are often used to collect information from large groups of people using scales that have been tested for validity and reliability. A researcher might be curious about how a supervisor sharing personal information with his or her subordinate affects way the subordinate perceives his or her supervisor. The researcher could create a survey where respondents answer questions about a) the information their supervisors self-disclose and b) their perceptions of their supervisors. The data collected about these two variables could offer interesting insights about this communication. As you would guess, an experiment would not work in this case because the researcher needs to assess a real relationship and they need insight into the mind of the respondent.

Content Analysis : Content analysis is used to count the number of occurrences of a phenomenon within a source of media (e.g., books, magazines, commercials, movies, etc.). For example, a researcher might be interested in finding out if people of certain races are underrepresented on television. They might explore this area of research by counting the number of times people of different races appear in prime time television and comparing that to the actual proportions in society.

Meta-Analysis : In this technique, a researcher takes a collection of quantitative studies and analyzes the data as a whole to get a better understanding of a communication phenomenon. For example, a researcher might be curious about how video games affect aggression. This researcher might find that many studies have been done on the topic, sometimes with conflicting results. In their meta-analysis, they could analyze the existing statistics as a whole to get a better understanding of the relationship between the two variables.

Qualitative research is interested in exploring subjects’ perceptions and understandings as they relate to communication. Imagine two researchers who want to understand student perceptions of the basic communication course at a university. The first researcher, a quantitative researcher, might measure absences to understand student perception. The second researcher, a qualitative researcher, might interview students to find out what they like and dislike about a course. The former is based on hard numbers, while the latter is based on human experience and perception.

Qualitative researchers employ a variety of different methodologies. Some of the most popular are interviews, focus groups, and participant observation. To better understand these research methods, you can explore the following examples:

Interviews : This typically consists of a researcher having a discussion with a participant based on questions developed by the researcher. For example, a researcher might be interested in how parents exert power over the lives of their children while the children are away at college. The researcher could spend time having conversations with college students about this topic, transcribe the conversations and then seek to find themes across the different discussions.

Focus Groups : A researcher using this method gathers a group of people with intimate knowledge of a communication phenomenon. For example, if a researcher wanted to understand the experience of couples who are childless by choice, he or she might choose to run a series of focus groups. This format is helpful because it allows participants to build on one another’s experiences, remembering information they may otherwise have forgotten. Focus groups also tend to produce useful information at a higher rate than interviews. That said, some issues are too sensitive for focus groups and lend themselves better to interviews.

Participant Observation : As the name indicates, this method involves the researcher watching participants in their natural environment. In some cases, the participants may not know they are being studied, as the researcher fully immerses his or herself as a member of the environment. To illustrate participant observation, imagine a researcher curious about how humor is used in healthcare. This researcher might immerse his or herself in a long-term care facility to observe how humor is used by healthcare workers interacting with patients.

Rhetorical research (or rhetorical criticism) is a form of textual analysis wherein the researcher systematically analyzes, interprets, and critiques the persuasive power of messages within a text. This takes on many forms, but all of them involve similar steps: selecting a text, choosing a rhetorical method, analyzing the text, and writing the criticism.

To illustrate, a researcher could be interested in how mass media portrays “good degrees” to prospective college students. To understand this communication, a rhetorical researcher could take 30 articles on the topic from the last year and write a rhetorical essay about the criteria used and the core message argued by the media.

Likewise, a researcher could be interested in how women in management roles are portrayed in television. They could select a group of popular shows and analyze that as the text. This might result in a rhetorical essay about the behaviors displayed by these women and what the text says about women in management roles.

As a final example, one might be interested in how persuasion is used by the president during the White House Correspondent’s Dinner. A researcher could select several recent presidents and write a rhetorical essay about their speeches and how they employed persuasion during their delivery.

Taking a mixed methods approach results in a research study that uses two or more techniques discussed above. Often, researchers will pair two methods together in the same study examining the same phenomenon. Other times, researchers will use qualitative methods to develop quantitative research, such as a researcher who uses a focus group to discuss the validity of a survey before it is finalized.

The benefit of mixed methods is that it offers a richer picture of a communication phenomenon by gathering data and information in multiple ways. If we explore some of the earlier examples, we can see how mixed methods might result in a better understanding of the communication being studied.

Example 1 : In surveys, we discussed a researcher interested in understanding how a supervisor sharing personal information with his or her subordinate affects the way the subordinate perceives his or her supervisor. While a survey could give us some insight into this communication, we could also add interviews with subordinates. Exploring their experiences intimately could give us a better understanding of how they navigate self-disclosure in a relationship based on power differences.

Example 2 : In content analysis, we discussed measuring representation of different races during prime time television. While we can count the appearances of members of different races and compare that to the composition of the general population, that doesn’t tell us anything about their portrayal. Adding rhetorical criticism, we could talk about how underrepresented groups are portrayed in either a positive or negative light, supporting or defying commonly held stereotypes.

Example 3 : In interviews, we saw a researcher who explored how power could be exerted by parents over their college-age children who are away at school. After determining the tactics used by parents, this interview study could have a phase two. In this phase, the researcher could develop scales to measure each tactic and then use those scales to understand how the tactics affect other communication constructs. One could argue, for example, that student anxiety would increase as a parent exerts greater power over that student. A researcher could conduct a hierarchical regression to see how each power tactic effects the levels of stress experienced by a student.

As you can see, each methodology has its own merits, and they often work well together. As students advance in their study of communication, it is worthwhile to learn various research methods. This allows them to study their interests in greater depth and breadth. Ultimately, they will be able to assemble stronger research studies and answer their questions about communication more effectively.

Note : For more information about research in the field of communication, check out our Guide to Communication Research and Scholarship .

research methods for communication

Media and Communication

  • Encyclopedias and Background Information
  • Finding Books and E-books
  • Finding Articles
  • Primary Sources

Communications Research Methods: Articles & Books

  • Methods in Media Studies (MiMS)

Getting help from NYU Library Data Services

For assistance, please submit a request .  You can also reach us via the chat below, email [email protected] , or join Discord server .

If you've met with us before,                        tell us how we're doing .

Service Desk and Chat

Bobst Library , 5th floor

Staffed Hours: Summer 2024

Mondays:  12pm - 5pm         Tuesdays:  12pm - 5pm         Wednesdays:  12pm - 5pm         Thursdays:  12pm - 5pm         Fridays:  12pm - 5pm        

Data Services closes for winter break at the end of the day on Friday, Dec. 22, 2023. We will reopen on Wednesday, Jan. 3, 2024.

Quantitative Data

Interested in using quantitative data? Need some assistance using SPSS?  Make a one-on-one appointment with the Data Service Studio by emailing [email protected].

For full details see the Data Service Studio page.

Woman in a t-shirt and scarf is sitting at a desk flipping through a magazine.

  • Communication Research Methods II: A Sourcebook An updated version of the "bible" of how to do research in communication and media studies.
  • Handbook of Media and Communication Research: Qualitative and Quantitative Methodologies A great overview of both qualitative and quantitative approaches to content analysis and media studies research.
  • Qualitative Media Analysis The authors of this work show readers how to obtain, categorize, and analyze different media documents. They look at traditional primary documents such as newspapers and magazines but also at more recent forms–television newscasts and cyberspace.
  • Mass Communications Research Methods Originally published in 1988. Step-by-step, this book leads students from problem identification, through the mazes of surveys, experimentation, historical/qualitative studies, statistical analysis, and computer data processing to the final submission and publication in scientific or popular publications.
  • Communication Research Methods in Postmodern Culture: A Revisionist Approach The second edition of Communication Research Methods in Postmodern Culture continues to explore research from a postmodern perspective. Typical qualitative and quantitative research methods are adjusted to fit the needs of contemporary culture.
  • << Previous: Primary Sources
  • Next: Course-Specific Guides >>
  • Last Updated: May 23, 2024 2:40 PM
  • URL:

Three Types of Communication Research Methods: Quantitative, Qualitative, and Participatory

  • Reference work entry
  • First Online: 09 June 2020
  • Cite this reference work entry

research methods for communication

  • Jan Servaes 2 , 3  

2300 Accesses

This chapter presents a brief overview of the three types of communication research methods being applied in development communication settings: quantitative, qualitative, and participatory. This chapter attempts to outline the relative characteristics and merits of these approaches to research and to emphasize some of the philosophical issues which underpin them. It discusses the strengths and weaknesses of each and highlights the benefits of a more normative approach focused on the “poor” in society.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
  • Available as EPUB and PDF
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Addo H (1985) Beyond Eurocentricity: transformation and transformational responsibility. In: Addo H, Amin S, Aseniero G et al (eds) Development as social transformation. Reflections on the global problematique. Hodder and Stoughton, London

Google Scholar  

Anderson J, Meyer T (1988) Mediated communication. A social action perspective. Sage, Newbury Park

Argyris C, Putnam R, Smith D (1985) Action science. Jossey-Bass, San Francisco

Berg B, Lune H (eds) (2014) Qualitative research methods for the social sciences. Pearson, Essex

Bhattacherjee A (2012) Social science research: principles, methods, and practices. Textbooks collection 3.

Bogaert MVD, Bhagat S, Bam NB (1981) Participatory evaluation of an adult education programme. In: Fernandes W, Tandon R (eds) Participatory research and evaluation: experiments in research as a process of liberation. Indian Social Institute, New Delhi

Bruhn Jensen H (ed) (2013) A handbook of media and communication research: qualitative and quantitative methodologies. Routledge, New York

Bryman A (1984) The debate about quantitative and qualitative research: a question of method or epistemology? Br J Sociol 35(1):75–92

Article   Google Scholar  

Bryman A, Burgess R (eds) (1994) Analyzing qualitative data, Routledge, London

Burke K (1968) Language as symbolic action. University of California Press, Berkeley

Chamarik S (1993) Democracy and development. A cultural perspective. Local Development Institute, Bangkok

Chambers R (2007) Ideas for development. Earthscan, London

Chambers R (2008) Revolutions in development inquiry. Earthscan, London

Chantana P, Wun Gaeo S (1985) Participatory research and rural development in Thailand. In: Farmer’s Assistance Board (ed) Participatory research: response to Asian people’s struggle for social transformation. Farmer’s Assistance Board, Manila

Christians C, Carey J (1981) The logic and aims of qualitative research. In: Stempel G, Weaver D, Westley B (eds) Research methods in mass communication. Prentice Hall, Englewood Cliffs

Cooke B, Kothari U (eds) (2001) Participation. The new tyranny? Zed Books, London

D’Abreo D (1981) Training for participatory evaluation. In: Fernandes, Tandon R (eds) Participatory research and evaluation: experiments in research as a process of liberation. Indian Social Institute, New Delhi

Denzin NK, Lincoln YS (eds) (2005) The Sage handbook of qualitative research, 3rd edn. Sage, Thousand Oaks

Dervin B (1982) Citizen access as an information equity issue. In: Schement JR, Gutierrez F, Sirbu M (eds) Telecommunications policy handbook. Praeger, New York

Dervin B, Huesca R (1999) The participatory communication for development narrative: an examination of meta-theoretic assumptions and their impacts. In: Jacobson T, Servaes J (eds) Theoretical approaches to participatory communication. IAMCR book series. Hampton Press, Creskill, pp 169–210

Ewen S (1983) The implications of empiricism. J Commun 33(3):219

Fals Borda O (1988) Knowledge and people’s power: lessons with peasants in Nicaragua, Mexico, and Colombia. Indian Social Institute, New Delhi

Fals Borda O, Rahman HA (eds) (1991) Action and knowledge: breaking the monopoly with participatory action research. Intermediate Technology Publications, London

Farrington J (1988) Farmer participatory research: editorial introduction. Exp Agric 24:269

Friberg M, Hettne B (1985) “The greening of the world”, development as social transformation. Westview, Boulder

Fuglesang A (1984) The myth of people’s ignorance. In: Development dialogue. The Dag Hammarskjold Foundation, Uppsala, pp 1–2

Geertz C (1973) The interpretation of cultures. Basic Books, New York

Halloran JD (1981) The context of mass communication research. In: Mcanany E, Schnitman J, Janus N (eds) Communication and social structure: critical studies in mass media research. Praeger, New York

Harris M (1980) Cultural Materialism: The Struggle for a Science of Culture, Vintage, New York

Harris M (2001) Cultural materialism: the struggle for a science of culture. Rowman & Littlefield, Lanham/Oxford

Huesca R (2003) Tracing the history of participatory communication approaches to development: a critical appraisal. In: Servaes J (ed) Approaches to development. Studies on communication for development. UNESCO, Paris

Huizer G (1989) Action research and People’s participation. An introduction and some case studies. Third World Centre, Nijmegen

Jacobson T (1993) A pragmatist account of participatory communication research for national development. Commun Theory 3(3):214–230

Jacobson T, Servaes J (eds) (1999) Theoretical approaches to participatory communication. IAMCR book series. Hampton Press, Creskill

Kennedy TW (1984) Beyond advocacy: an animative approach to public participation. Dissertation, Cornell University

Kennedy TW (2008) Where the rivers meet the sky: a collaborative approach to participatory development. Southbound, Penang

Kronenburg J (1986) Empowerment of the poor. A comparative analysis of two development endeavours in Kenya. Third World Center, Nijmegen

Kuhn T (1962) The structure of scientific revolutions, 2nd edn, enlarged. University of Chicago, Chicago

Lennie J, Tacchi J (2013) Evaluating communication for development. A framework for social change. Routledge Earthscan, London

Book   Google Scholar  

Lie R (2003) Spaces of intercultural communication. An interdisciplinary introduction to communication, culture, and globalizing/localizing identities. IAMCR book series. Hampton Press, Creskill

Mckee N, Manoncourt E, Saik YC, Carnegie R (eds) (2000) Involving people, evolving behaviour. Southbound, Penang

Midgley J (ed) (1986) Community participation, social development, and the state. Methuen, London

Plaisance PL (2011) Moral agency in media: toward a model to explore key components of ethical practice. J Mass Media Ethics 26(2):96–113

PPDHRRA – Philippine Partnership for the Development of Human Resources in Rural Areas (1986) Participatory research guidebook. Philippine Partnership for the Development of Human Resources in Rural Areas, Laguna

Robson C (1995) Real world research. A resource for social scientists and practitioner-researchers. Blackwell, Oxford

Rockhill K (Fall, 1982) Researching participation in adult education: the potential of the qualitative perspective. Adult Educ 33(1):3

Servaes J (1999) Communication for development. One world, multiple cultures. Hampton Press, Creskill

Servaes J (2012) Homo academicus: Quo vadis? In: Silvia N-Z, Karyn H (eds) Global academe: engaging intellectual discourse. Palgrave Macmillan, New York, pp 85–98. 230pp

Chapter   Google Scholar  

Servaes J, Jacobson T, White S (eds) (1996) Participatory communication for social change. Sage, New Delhi

Tacchi J, Slater D, Hearn G (2003) Ethnographic action research. UNESCO, Paris

Tandon R (1981) Participatory evaluation and research: main concepts and issues. In: Fernandes W, Tandon R (eds) Participatory research and evaluation: experiments in research as a process of liberation. Indian Social Institute, New Delhi

Tandon R (1985) Participatory research: issues and prospects. In: Farmer’s Assistance Board (ed) Response to Asian people’s struggle for social transformation. Farmer’s Assistance Board, Manila

Tandon R (2002) Participatory research: revisiting the roots. Mosaic Books, New Delhi

Thayer L (1983) On ‘doing’ research and ‘explaining’ things. In: Gerbner G (ed) Ferment in the Field. J Commun 33(4)

Tuhiwai Smith L (1999) Decolonizing methodologies. Research and indigenous peoples. Zed Books, London

Van Hemelrijck A (2013) Powerful beyond measure? Measuring complex systemic change in collaborative settings. In: Servaes J (ed) Sustainability, participation and culture in communication. Theory and praxis. Intellect-University of Chicago Press, Bristol/Chicago

White R (1982) Contradictions in contemporary policies for democratic communication, Paper IAMCR conference, Paris, September

White R (1984) The need for new strategies of research on the democratization of communication, Paper ICA conference, San Francisco, May

Whyte WF (ed) (1989) Learning from the field. A guide from experience. Sage, Beverly Hills

Whyte WF (ed) (1991) Participatory action research. Sage, Newbury Park

Download references

Author information

Authors and affiliations.

Department of Media and Communication, City University of Hong Kong, Hong Kong, Kowloon, Hong Kong

Jan Servaes

Katholieke Universiteit Leuven, Leuven, Belgium

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Jan Servaes .

Editor information

Editors and affiliations, rights and permissions.

Reprints and permissions

Copyright information

© 2020 Springer Nature Singapore Pte Ltd.

About this entry

Cite this entry.

Servaes, J. (2020). Three Types of Communication Research Methods: Quantitative, Qualitative, and Participatory. In: Servaes, J. (eds) Handbook of Communication for Development and Social Change. Springer, Singapore.

Download citation


Published : 09 June 2020

Publisher Name : Springer, Singapore

Print ISBN : 978-981-15-2013-6

Online ISBN : 978-981-15-2014-3

eBook Packages : Literature, Cultural and Media Studies Reference Module Humanities and Social Sciences Reference Module Humanities

Share this entry

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

  • Publish with us

Policies and ethics

  • Find a journal
  • Track your research
  • Tools and Resources
  • Customer Services
  • Communication and Culture
  • Communication and Social Change
  • Communication and Technology
  • Communication Theory
  • Critical/Cultural Studies
  • Gender (Gay, Lesbian, Bisexual and Transgender Studies)
  • Health and Risk Communication
  • Intergroup Communication
  • International/Global Communication
  • Interpersonal Communication
  • Journalism Studies
  • Language and Social Interaction
  • Mass Communication
  • Media and Communication Policy
  • Organizational Communication
  • Political Communication
  • Rhetorical Theory
  • Share This Facebook LinkedIn Twitter

Article contents

Methods for intercultural communication research.

  • John Oetzel , John Oetzel Waikato Management School, The University of Waikato
  • Saumya Pant Saumya Pant Mudra Institute of Communications, Ahmedabad (MICA)
  •  and  Nagesh Rao Nagesh Rao Partner, Siya Consulting
  • Published online: 09 May 2016

Research on intercultural communication is conducted using primarily three different methodological approaches: social scientific, interpretive, and critical. Each of these approaches reflects different philosophical assumptions about the world and how we come to know it. Social scientific methods often involve quantitative data collection and research approaches such as surveys and experiments. From this perspective, intercultural communication is seen as patterns of interaction, and we seek to explain and understand these patterns through clear measurement and identification of key independent variables. Interpretive methods often involve qualitative data collection and research approaches such as interviews and ethnographic observation. From this perspective, intercultural communication and meaning is created through interaction, and we seek to understand these meanings by exploring the perspectives of people who participate as members of cultural communities. Critical methods often involve qualitative data collection and research approaches such as interviews and textual critique. From this perspective, intercultural communication involves inequalities that can be attributed to power and distortions created from (mis)use of this power. Critical scholars seek to unmask domination and inequality. Most scholars utilize one of these primary approaches given the consistency with their world views, theories, and research training. However, there are creative possibilities for combining these approaches that have potential for fuller understanding of intercultural communication.

  • social science methods
  • interpretive methods
  • critical methods
  • quantitative
  • qualitative
  • intercultural communication


Our worldview shapes what is “interesting” to a particular audience, what is considered a problem, what problem is interesting to study, and whether the goal of studying a problem is to analyze the problem, to analyze and solve the problem, or to analyze, solve, and implement the solution. Our worldview defines if an issue is a problem or not and if we need to come up with a solution. For example, behaviors associated with attention deficit hyperactivity disorder (ADHD) are seen as a problem in the United States, and there are medications to solve the problem. In India, the same set of behaviors among children is seen as what children tend to do, as normal and not as a problem.

Our worldview not only shapes what we see as an interesting problem to study but also the methodology we use to study the problem. The purpose of this article is to describe, and explore integration of, the three main methodological perspectives in studying intercultural communication issues: social scientific, interpretive, and critical. First, the ontological, epistemological, and axiological assumptions underlying each of these methodological perspectives are explored. Then, for each methodological perspective, common methods and types of data collected and some exemplars are identified. Finally, we offer traditional integration of the three approaches and also alternate methodological perspectives to study intercultural issues from a non-Western lens.

Ontology, Epistemology, and Axiology

Ontology is the study of the researcher’s orientation to reality. In the social scientific perspective, the researcher views the world objectively in that there is a world outside of us that can be systematically studied. Researchers from this perspective use a deductive approach and are keen to explain and predict phenomena. Social scientific ontology provides clarity and direction due to its rigorous questioning of plausibility and reduction of subjectivity. In contrast and as a reaction to the social scientific perspective, interpretive researchers argue that the observer and the observed are subjective and the most important lessons are in how they co-create meaning. If the social scientists take a deterministic view of human behavior, interpretivists thrive in a person’s free will. Critical theorists focus particularly on social injustices and inequalities in life. Researchers in this area explore how social structures create power inequalities and injustices. Thus, they believe that power differences are at the base of social transactions (Scotland, 2012 ). Any ontological investigation for a critical theorist will thus have to help unearth these inequities.

Epistemology looks at how we come to know a chosen phenomenon and thus how researchers study this phenomenon. Social scientists, interested in assessing objective reality (or at least reduced subjectivity), use a scientific method to collect empirical evidence. They focus particularly on causal relationships between phenomena and generally use quantitative approaches to collect data. The basis of their assessment and data collection is the premise that objects have an existence independent of the knower (Cohen et al., 2007 ). Interpretivists, who are interested in situational and contextual meaning, generally use qualitative methods to assess participants’ sense of reality. They are not exploring one truth, but the play of multiple truths simultaneously. They do so by studying individual interactions and the historical and cultural contexts in which these individuals interact. Critical researchers use a variety of qualitative methods to explore, for example, how language is used to create power imbalances or how mass media is used to avoid critical thinking. Critical scholars are particularly sensitive to the overdependence on empirical and social scientific evidence. They do so as critical investigations are premised on the fact that social/positional power determines what is considered knowledge (Cohen et al., 2007 ).

Axiology explores the values that guide a researcher’s questions, the methods used to collect and analyze data, the interpretation of the data, and the implications of the findings. Social scientists study phenomena to find the truth, which, in turn, guides specific types of action. They are focused on exploring what is referred to as the value axiom, or how much a phenomenon being studied fulfills the requirements of the concept to which it belongs (Kelleher, 2013 ). Both interpretivists and critical theorists are interested in describing what exists, how the participants in the community interpret phenomena, with critical theorists particularly interested in reducing class imbalances and other forms of oppression. Interpretivists are axiologically determined to encourage the fact that observations drawn can always be disagreed upon and reopened to interpretation. With respect to control, social scientists wish to control as many variables as possible, narrowing down the causal pattern to the variables under study. Interpretivists seek active participation in the study to understand how they view reality. Critical theorists are particularly aware of the community members’ need to take control of their own situations. With this brief overview in mind, we now explain the methodological approaches of the social scientific, interpretive, and critical perspectives; the types of data collected; some exemplars for each perspective; and some general concerns about each of the methods.

Social Science Methods

Social science research methods address questions related to both cross-cultural and intercultural communication. Much of the foundational work on intercultural communication research is based on comparisons of two or more cultures. Both forms of communication research try to enhance the comprehension of communication that are mediated by and through cultural context (Sponcil & Gitimu, 2010 ). These comparisons helped to identify how the normative and subjective aspects of culture vary across cultures and presumably provided information about what to expect when interacting with members from different cultures. This type of research is classified as cross-cultural. In contrast, intercultural communication is the exchange of messages between people from different cultural groups (Gudykunst, 2003a ). Regardless of the interest in cross-cultural or intercultural communication, the social scientific perspective seeks to understand and predict the effect of culture on communication variables and the subsequent effect of communication on various outcomes. Thus, the methods of study are similar. This section reviews the three most prominent social scientific methods providing an example of each. Additionally, the types of data generated and methodological concerns are discussed.

There are three methods used by most social scientific researchers to study cross-cultural and intercultural research: (a) survey questionnaire, (b) experimental design, and (c) content analysis. The survey questionnaire is by far the most frequently used research method (e.g., Oetzel & Ting-Toomey, 2003 ; Rao, Singhal, Ren, & Zhang, 2001 ). It is typically a self-administered and self-report instrument that is distributed to large samples in multiple cultures. Most cross-cultural comparisons utilize self-report questionnaires because of the difficulty of collecting data from large samples in multiple cultures using other methods. Finally, self-report questionnaires are relatively easy to construct. Numerous cross-culturally valid scales exist, and methodological difficulties have been clearly identified (Gudykunst, 2003b ). While not easy to overcome, methodological difficulties of survey questionnaires are manageable (see below for more detail). Survey questionnaires provide detailed description of cultural associations of communication behavior and outcomes and allow for comparisons to other cultures.

Hanasono, Chen, and Wilson’s ( 2014 ) study of perceived discrimination, social support, and coping among racial minority university students is an example of survey research. The authors surveyed 345 students, half international students and half U.S. students, about their acculturation, experiences with discrimination, support, and coping needs. They found that the level of acculturation helped to explain students’ need for support and how they coped with discrimination.

Experimental designs are highly regarded social scientific research because of the control of variables, which enables causal relationships to be examined. Culture is not a variable that lends itself well to experimental manipulation, and thus experimental designs are relatively rare in this line of research. Rather than experimental controlling culture, researchers typically use quasi-experimental designs manipulating the composition of groups or dyads to be intra- or intercultural (e.g., Cai, Wilson, & Drake, 2000 ; Oetzel, 1998 ). These experiments collect a combination of self-report information (e.g., cultural and individual variables) as well as videotaped interaction. Additionally, some researchers have used experimental conditions on survey questionnaires (e.g., Han & Cai, 2010 ). These studies utilize stimulus variables (e.g., contextual features) that ask participants to respond to specific situations.

Brinson and Stohl’s ( 2012 ) study of media framing on attitudes toward Muslims, civil liberties, and counterterrorism policies is an example of experimental design. They used a Solomon four-group design involving 371 U.S. adults to compare the media framing of “domestic homegrown” and “international” terrorism of the London bombings in 2005 . The authors used video segments from actual broadcasts on July 7, 2005 , and edited them together to create an approximately 10-minute video for each of the two conditions. The authors found that media frames of homegrown terrorism produced greater fear than the international framing. Fear resulted in greater support for restricting civil liberties of Muslims and, under certain conditions, general negative feelings toward Muslims.

A third method used in social scientific research is content analysis of media sources. This method is utilized to identify patterns prevalent in the media (e.g., Dixon & Azocar, 2006 ; Klein & Shiffman, 2006 ). Additionally, some researchers survey participants for their reactions about media patterns. Content analysis, while time consuming, is convenient and inexpensive since the only access needed is a recording or transcript of the artifact of study. It involves the use of a coding scheme to provide an “objective” description of the media and thus insights into cultural values and behaviors. The categorizations are then compared across cultures. When these categorizations are compared, it is done on the basis of frames, which are defined as a “schema of interpretation, collection of anecdotes, and stereotypes” (Cissel, 2014 , p. 67). Once these frames are determined, the way in which individuals deal with their realities within and across cultures can be studied.

An example of such content analysis was the study of the coverage of the Fukushima nuclear accident in Japan in two Belgian newspapers: Le Soir and De Standard (Perko et al., 2011 ). The time period of the study was from March 11, 2011, to May 11, 2011 . Every article was coded by two independent coders. The authors had begun their study with a question as to how the framing of the question of nuclear power would appear in the two Belgian newspapers. They arrived at the conclusion that the reporting was mostly neutral. Further, since the Fukushima nuclear accident was in a country quite remote, the articles did not frame the issue as an example of a possible threat to their own country from nuclear power plants.

Data Analysis and Methodological Concerns

Data from these three methods are quantified to allow for statistical analysis. All forms of data must be reduced to categories that are independent from one another (exhaustive and exclusive categories). These can include frequency counts of behaviors, sequence of behaviors, and self-report information on numerical scales. Data are then analyzed with statistical software to determine associations between cultural (independent) and communication (dependent) variables (outcomes are dependent variables with culture and/or communication as independent). The nature of analysis depends on the numerical measurement of the variables, but frequent tests include t -tests, analysis of variance, correlation, and regression. Additionally, complex modeling of dependent variables can be undertaken using, for example, structural equation modeling and hierarchical linear modeling. The key concern with the statistical tests is accounting for variance in the dependent variables. The more variance explained means the “more important” a cultural factor is for communication behavior. Because of the vast number of factors that explain human behavior, intercultural researchers believe that as little as 5–10% of variance explained is meaningful.

There are four concerns for data analysis in social scientific research: (a) reliability, (b) measurement validity, (c) internal validity, and (d) external validity. Reliability is reproducibility. For the aforementioned methods, two types of reliability are relevant. First, internal consistency of measures is usually measured with Cronbach’s alpha. Second, when completing content or interaction analysis, intercoder reliability (agreement between two or more coders) is important and measured with Cohen’s K or Scott’s pi (or the like). Reliability means a researcher has consistent measures, whereas validity focuses on accurate information.

Validity is a combination of measurement, internal, and external validity (depending on the goals in the study). Measurement validity focuses on the accuracy with which a scale (or coding scheme) is measuring what is supposed to be measured. Internal validity is the strength to which a researcher can conclude that the independent variable is associated with the dependent variable as hypothesized. Internal validity is established by eliminating rival explanations for statistical associations through statistical or experimental control of confounding (or nuisance) variables. External validity is the degree to which a study’s results can be generalized to the larger populations from which a sample was drawn. In intercultural research, researchers are more concerned with measurement and internal validity than external validity.

While these general methodological concerns are true for all social science research, there are also unique concerns with cross-cultural/intercultural communication research (Gudykunst, 2003b ; Levine, Park, & Kim, 2007 ). Gudykunst ( 2003b ) outlined a number of concerns with cross-cultural research, but chief among the methodological issues is establishing equivalence. In order to make cross-cultural comparisons (and have valid measures for intercultural research), researchers need to ensure that the constructs and measures are equivalent on five levels. First, constructs must be functionally equivalent; that is, the construct must work the same way in the cultures under study. Second, constructs must be conceptually equivalent; that is, the construct must have the same meaning within the cognitive system of the members of cultures being examined. Third, linguistic equivalence for constructs refers having language that is equivalent. Linguistic equivalence is often established through translating and backtranslating of measures. Fourth, metric equivalence is established by ensuring that participants in different cultures do not respond to numerical scales in different ways (e.g., one cultural group may not use the extreme scores in a scale). Finally, researchers need to take care and establish that there is sample equivalence in the two cultural groups. The samples need to be comparable (e.g., similar age, gender, education, etc.). Fletcher and colleagues ( 2014 ) explore the steps needed to statistically ensure equivalence in measurement across multiple cultures. Establishing equivalence on these issues helps to eliminate rival explanations and further ensures that differences found are due to cultural differences. In addition to such methodological rigor, scholars from other orientations argue that it is also imperative for the researcher to be reflexive and aware of theoretical and methodological centeredness that can come from such systematic rigor (Asante, Miike, & Yin, 2008 ).

Interpretive Methods

Interpretive scholars are interested in unearthing multiple simultaneous truths, believe in a person’s free will, acknowledge that the known and the knower cannot be separated, and believe that interpretation is based on one’s persuasive abilities. Striving for meaning, interpretive scholars generally use a variety of qualitative methods to study specific intercultural phenomena. As a result of this, interpretivists examine theoretical limits by comparing results from multiple forms of research about the same phenomenon (Szabo, 2007 ). For this article, we focus on ethnography of communication and interpretive interviews as these are two common approaches. We then discuss the general methodological issues in collecting and analyzing interpretive data.

Ethnography of communication (EOC) is a method to study the relationship between language and culture through extensive field experience. The concept of the ethnography of communication was developed by Dell Hymes (Hall, 2002 ). It can be defined as the discovery and explication of the rules for contextually appropriate behavior in a community or group or what the individual needs to know to be a functional member of the community. EOC applies ethnographic methods to understand the communication patterns of a speech community (Philipsen, 1975 ). A speech community is a group of speakers who share common speech codes and use these codes based on a specific situation. From the presence or absence of certain speech codes, one can interpret the culture of a community with its shared values, beliefs, and attitudes. In his classic study, Philipsen ( 1975 ) explored the communication patterns of white males in a predominantly blue-collar neighborhood called “Teamsterville” in South Chicago. Philipsen lived in the community for several years and worked and interacted as a member of the community while also conducting his research. Results from this study explained when talk was appropriate, at what levels, and when action was more appropriate than talk. When two men were of similar backgrounds, of more or less equal status, and were close friends, they could talk to each other. There was less talk when the relationship was asymmetrical (e.g., father–son and husband–wife). The least amount of talk occurred when a “Teamsterville” was responding to an insult or trying to assert his power over someone. It is in these instances that action was more appropriate than words. If a man did talk during this interaction, he was seen as not masculine enough. Another interesting study was conducted by Radford et al. ( 2011 ). The study focused on applying EOC to the case of virtual reference context. Here, the researchers focused on the interactions that constitute the context in which the participants make verbal statements and coordinate them with other statements in order to closely analyze the relational barriers and relational facilitators. The interactions spanned a 23-month time period ( July 2004–May 2006 ), and the transcripts of 746 live chats of this period were studied. The researchers were able to conclude from their study that when professional librarians chatted, they were more formal, less free with accepted online abbreviations, whereas students were more comfortable with using abbreviations and other turns of phrases. One of the conclusions the researchers drew was that if the librarians used more informal language they would appear more friendly and approachable.

Interpretive interviews are a second common approach. The purpose of the interpretive interview is to uncover insider meanings and understandings from the perspective of the participants (Denzin, 2001 ). According to Denzin ( 2001 ), the characteristic of interpretive interviews is that they allow us to understand the society in which we live, which is referred to as an interview society. Typically, these interviews are one-on-one and face-to-face interviews designed to elicit in-depth information. The interviews can focus on narratives, topics, perspectives, and opinions and often are conducted in a semi-structured manner (although unstructured interviews are sometimes conducted). One of the reasons why the semi-structured and/or narrative form is used is to allow for deeper and embedded meanings that might elude a more inquiry-based approach. An example of interpretive interviews is Baig, Ting-Toomey, and Dorjee’s ( 2014 ) study of meaning construction of the South Asian Indian term izzat (face) in intergenerational contexts. The authors interviewed six younger (aged 31–40) and six older (aged 55–72) South Asian Indian American women about face concerns in their intergenerational family communication situations. The authors found that family izzat is of primary importance in these contexts and that the motif of respect is central to the meaning of izzat . They also identified differences in the younger and older facework strategies.

The primary focus of analyzing interpretive research data is rather nicely summarized by Carbaugh ( 2007 ):

It is important to emphasize the interpretive task before the analyst: while engaging in a communication practice, an analyst seeks to understand what range of meanings is active in that practice, when it is getting done. The analyst sets out to interpret this practice, what is being presumed by participants for it to be what it is, that is, to understand the meta-cultural commentary imminent in it. What all does this practice have to say? (p. 174)

Thus, the interpretive scholar analyzes data in order to describe and interpret.

Carbaugh ( 2007 ) identified two concerns in analyzing interpretive data—the framework used to analyze the semantic content of cultural discourse and the vocabulary used to formulate these contents. A researcher’s analysis of the content of the communication exchange also includes a meta-analysis of the subject, the object, the context, the history, and the stories revolving around the exchange. Carbaugh ( 2007 ) noted that “these cultural meanings—about personhood, relationships, action, emotion, and dwelling, respectively—are formulated in cultural discourse analyses as radiants of cultural meaning” (p. 174). These radiants of cultural meanings focus on personhood and identity, relating and relationships, meanings about acting, action and practice, meanings about emotions and feelings, and meanings about place or context.

Reliability and validity are explicated differently in interpretive research compared to social science research. If social scientific scholars are interested in consistency for reliability, interpretive scholars see reliability as the quality of the information obtained; does the data give us a richer, clearer understanding of the phenomena (Golafshani, 2003 )? Lincoln and Guba ( 1985 ) used the term “dependability” in place of reliability to assess the quality of a research project. For validity, it is important to assess the quality based on the specific paradigm used to conduct the qualitative research. Further, while many scholars argue that validity is not a critical concept for interpretive research, Lincoln and Guba ( 1985 ) explained that the “trustworthiness” of the data is similar to validity in social science research. Do the community of scholars conducting interpretive research view the data as meaningful, useful, and following the research protocols appropriately?

After having considered these general considerations, we now consider three specific data analysis approaches using in interpretive intercultural communication research including grounded theory, constant comparative analysis, and thematic analysis. Other data analytic approaches for data analysis include narrative analysis, conversational analysis, EOC, and interpretive phenomenological analysis. Grounded theory is a continuum of practices that are inductive and iterative aimed at recognizing categories and concepts in texts in order to integrate them to formal theoretical models (Corbin & Strauss, 2008 ). They begin with the observations, experiences, and stories, and through a process of coding, analysts identify a theoretical model to fit the data. Another important approach that interpretive scholars use is that of constant comparative analysis (CCA). CCA has often been used as a part of grounded theory, but it is now being used separately to analyze cross-cultural and intercultural communication. CCA is used to balance the etic perspective (participant as outsider) with the emic perspective (participant as insider) to ensure balance between cultural readings and theoretical frameworks. CCA ensures that all data in the relevant set are compared with all other data in the same set to make sure that no data are dismissed on thematic grounds (O’Connor et al., 2008 ). Further, CCA tries to accommodate the most relevant theories though they may appear disparate. A final prominent approach is thematic analysis. Thematic analysis is a flexible and yet rigorous approach of identifying and analyzing patterns or themes of meaning from data. Braun and Clarke ( 2006 ) identify a six-step process for conducting thematic analysis.

Critical Methods

From the critical perspective, relationships between cultural groups are often characterized by dominance and resistance. Communication between groups is based on certain understanding of culture and ethnicity that is fixed, reified, and essentialized and is informed by certain cultural assumptions that tend to be rooted in Euro-American traditions and worldviews (Asante et al., 2008 ). Hermans and Kempen ( 1998 ) argued that dominant approaches to knowledge favor static conceptualizations of culture. It is the creation of these static categories in which the Western understanding of the rest of the world dominates the intercultural relations that results in the reification of culturally homogeneous “ethnic” and racial groups. Consequently, this orientation undermines ways in which the self is understood in different cultures.

Critical and feminist scholars have consistently raised questions about power imbalance between researchers and researched in the field, suggesting that if researchers fail to explore how their personal, professional, and structural positions frame social scientific investigations, researchers may inevitably reproduce dominant gender, race, and class biases (Fairclough, 1995 ; Lazar, 2005 ). This section illustrates postcolonial ethnography and critical discourse analysis as approaches for intercultural discovery from the critical lens. Additionally, we introduce the role of self-reflexivity and consciousness-raising in the context of methodological concerns from the critical perspective.

A variety of approaches to critical issues exist such as critical race theory, decolonizing and indigenous methodologies, engaged methodologies, and performative methodologies (Willink, Gutierrez-Perez, Shukri, & Stein, 2014 ). In this article, we explore two prominent methods to illustrate some of the key elements to critical approaches given that we cannot cover all of them: postcolonial ethnography and critical discourse analysis.

Postcolonial ethnography seeks to disrupt and restructure established academic practices and modes of knowledge development and dissemination (Pathak, 2010 ). It attempt to do this by pointing out that gender roles, academic institutions, racial binaries, and other power structures are not apolitical. Postcolonial ethnography seeks to question the reification and valorization of supposed objective, scientific, and disembodied knowledge formations. Instead they seek to find alternate and embodied knowledge forms that accommodate the subjective and the personal.

While postcolonial and third world feminist scholars point to myriad ways in which relations of domination infuse ethnography, they also offer some guidance for negotiating power inherent in the practice of fieldwork (Spivak, 1999 ). This guidance takes the form of feminist geopolitics, which involves not only questioning hegemonic structures and dominant power structures but also offering alternatives to those structures (Koopman, 2011 ). Postcolonial scholars argue that the practice of ethnography among marginalized groups is historically tainted by ethnocentric biases in traditional ethnographic practice and research (Collins, 1990 ). Further, as philosopher Sandra Harding ( 1998 ) emphasized, ethnocentricism is structured into the institutional and academic practices so as to produce relationships oppressive to indigenous cultures in the so-called first world as well as third world countries.

An example of postcolonial inquiry is that of an ethnographic encounter (Irani et al., 2010 ). As a part of this inquiry, the company that the researchers studied, Ddesign, had to develop prototypical home water purification filters (Irani et al., 2010 ). The site of their study was various villages in India where they were supposed to study the feasibility of home water purifiers among the economically deprived households of the villages. The researchers later were told that when Ddesign first started their study, they had notions of privations in the lives of the householders. During their study, they found that the reality was quite different from their preconceptions. They realized that the definitions of privations that the company personnel had were not applicable to the people or to their living conditions. In fact, the researchers were told by the company personnel that the villagers had a very different worldview from that of the personnel. Thus, the researchers and the company personnel realized that one group’s notions of well-being and happiness were not necessarily applicable to another group no matter how universal those notions might be.

A second approach is critical discourse analysis (CDA). The creative combining of different approaches of lived experience, texts or discourses, and the social and political structures of power has resulted in popularity of cultural studies as a critical site for different modes of enquiry. According to Fairclough ( 1995 ), “many analysts are becoming increasingly hesitant in their use of basic theoretical concepts such as power, ideology, class, and even truth/falsity” (p. 15). In recent social scientific research, there has been a turn toward language or, more specifically, toward discourse. According to the feminist critical scholar Michelle M. Lazar ( 2005 ), discourse is a “site of struggle, where forces of social (re)production and contestation are played out” (p. 4). Critical discourse analysis is known for its overtly political stance and deals with all forms of social inequality and injustice. It includes the study of processes premised on the acts and discursive interactions of individuals and groups on which both the local and international contexts bring to bear their limits in the production of legislation, news making, and other such products of discursive interactions (van Dijk, 2008 ).

An example of critical discourse analysis in intercultural communication research is Chen, Simmons, and Kang’s ( 2015 ) study of identity construction of college students. The authors contextualize their study in an era of “postracial” utopia resulting during the Obama administration. They coin the term “Multicultural/multiracial Obama-ism (MMO)” to reflect this era and the prominent frame of colorblindness and multiculturalism prominent in media discourse. They examined 65 student essays about three cultural identities that stood out in a particular context. They analyzed the essays using CDA and found three frames that support this construction of postracial utopia: meritocracy, identity as self-chosen, and equality of opportunity despite privilege. They critique these frames and identify implications for teaching about intercultural communication and identity in the classroom.

Key methodological issues in the critical approach are the role of reflexivity, consciousness-raising, and limitations/possibilities of the reflective approach. A sociology-of-knowledge approach to critical scholarship reveals the role of reflexivity as a source of insight (Cook & Fonow, 1984 ). Reflexivity means the tendency of critical scholars to reflect upon, examine critically, and explore analytically the nature of the research process. To some extent, this tendency toward reflection is part of a tradition of attention to what Kaplan ( 1964 ) referred to as “logic-in-use” or the actual occurrences that arise in the inquiry, idealized and unreconstructed. Feminist and critical epistemology carries this tradition of reflection further by using it to gain insight into the assumptions about gender and intercultural relations underlying the conduct of inquiry. This is often accomplished by a thoroughgoing review of the research setting and its participants, including an exploration of the investigator’s reactions to doing the research.

One of the ways in which reflexivity is employed involves the concept of consciousness-raising, a process of self-awareness familiar to those involved with the women’s movement. Underlying much of the reflexivity found in feminist scholarship is the notion found in the earlier work of scholars such as W. E. B. DuBois ( 1969 ) and Paulo Freire ( 1970 ) that consciousness of oppression can lead to a creative insight that is generated by experiencing contradictions. Under ideal circumstances, transformation occurs, during which something hidden is revealed about the formerly taken-for-granted aspects of intercultural relations.

Consciousness-raising is employed in various ways by the critical scholar. The first way is through attention to the consciousness-raising effects of research on the researcher. Consciousness-raising is also involved in discussions of ways in which the research process influences subjects of the inquiry. Some authors view the research act as an explicit attempt to reduce the distance between the researcher and subjects (Collins, 1990 ). These approaches have provided critical and feminist researchers with a way to tap collective consciousness as a source of data and have provided participants in the research process with a way to confirm the experiences that have often been denied as real in the past. The applications of critical consciousness-raising and reflexivity can be seen in discourses surrounding terrorism and counterterrorism. This application can be seen in the study by Schmid ( 2013 ) about radicalization, deradicalization, and counter-radicalization. Schmid has observed that the usual causes such as poverty, social inequality, oppression, and neglect by the West have not been empirically tested satisfactorily, yet they are believed to be the primary causes of radicalization. The study provides three levels of analysis that can be used to understand how “radicals” are born and how that complex construction can be interrogated: the micro level, dealing with the individual level in terms of identity and self-reflection; the meso level, which deals with the socio-political milieu surrounding the individual; and the macro level, which refers to the larger society and governance that affect the individual. Further, these three levels of analysis can also be used to see how the continuum from radical to political undesirable and terrorist can be studied.

Finally, there are limitations and possibilities of reflective practice. Critical researchers use self-reflection about power as a tool to deepen ethnographic analysis and to highlight the dilemmas in fieldwork. The call for reflective practice has also been informed by critiques of postcolonial theorists who argue for self-reflexive understanding of the epistemological investments that shape the politics of method (Mohanty, 1991 ). Cultural studies scholars have also questioned the call to reflective practice, arguing that taken to the extreme, “constant reflexivity” can make “social interaction extremely cumbersome” (Hurtado, 1996 , p. 29). In contrast, the call to “accountability” is said to offer a more collective approach than the “individual self assessment of one’s perspective” that the term “reflexivity implies” (Hurtado, 1996 , p. 29). However, from point of ethnographic practice, it is seldom clear to whom one should be “accountable,” and therefore the term reflective practice seems to be appropriate.

Reflective practice indicates both individual self-assessment and collective assessment of research strategies. Hurtado ( 1996 ) emphasized that a “reflexive mechanism for understanding how we are all involved in the dirty process of racializing and gendering others, limiting who they are and who they can become” (p. 124) is a necessary strategy to help dismantle domination. Such reflective strategies can also help ethnographers bring to the surface “their own privilege and possible bias” as well as “addressing the difference between different constituencies” (Hurtado, 1996 , p. 160) within the communities they study.

Integrating Social Science, Interpretive, and Critical Research Methods

Each set of methods presented in this chapter has strengths and limitations. They address specific purposes that collectively are all important for the field of intercultural communication. Moreover, integrating the research methods provides richer insights than using any method by itself. However, these integrations still may have limitations in exploring non-Western contexts. Thus, this section explores integrations of the methods and alternative methods for intercultural inquiry.

Integrations of Methods

The integration of research methods involves using different types of methods at different phases (Cresswell & Plano Clark, 2011 ). In this manner, the methods are used one after another (or concurrently) depending on the research question associated with the larger research program. Four phasic designs are most prevalent: a) qualitative/interpretive methods used to create a quantitative (social science measure); b) qualitative (interpretive and critical) methods used to embellish quantitative findings (Big Quant, Little Qual); c) quantitative methods used to embellish qualitative findings (Big Qual, Little Quant); and d) social science, interpretive, critical methods used conjointly. Space limitations prohibit us from providing examples of all of these approaches, so we detail two of them.

Zhang and colleagues (Zhang & Oetzel, 2006 ; Zhang, Oetzel, Gao, Wilcox, & Takai, 2007 ) provide an example of how to create a cross-culturally valid measure of a construct. Their purpose was to measure teacher immediacy. Teacher immediacy is the psychology closeness that is communicated from a teacher to a student. There exist different measures of immediacy, but Zhang and Oetzel ( 2006 ) argued that prior Western measures were not applicable in Chinese classrooms (i.e., they did not have conceptual equivalence). To address this issue, they first conducted open-ended interviews with Chinese students to identify themes associated with the meanings of immediacy. This phase of the research involved interpretive research methods as they put primacy on emic meanings. In the second phase, they used the emic meanings to create an operational measure of three dimensions of teacher immediacy (instructional, relational, and personal). This measure was administered to college students, and the data were analyzed with confirmatory factor analysis. The results dimensions were found to be internally consistent and had construct validity as they correlated with existing scales in expected directions. Zhang et al. ( 2007 ) then continued the development of the scale by administering the scale to college students in four national cultures: China, Japan, Germany, and the United States. With these data, the authors used confirmatory factor analysis to see if the three-dimensional model of teacher immediacy held up in each culture. They found cross-cultural support for the model and also the construct validity of the scales. Thus, their thorough testing from the interpretive phase to the social scientific phase led to the development of a teacher immediacy scale that has valid dimensions in at least four national cultures.

An example of integrating critical, social scientific, and interpretive methods into the same research program can be seen in the work on whiteness ideology (Nakayama & Martin, 1999 ). The project culminated in an edited book that included chapters using the various research methods. Whiteness ideology is the worldview that certain groups have privilege over others. It is labeled whiteness because whites tend to be the privileged groups in most societies. This research group’s work primarily focused on ethnic groups in the United States, but some international contexts were examined and other scholars have since examined international contexts we well (e.g., Collier, 2005 ). One part of the project examined the labels that white people in the United States prefer through a survey (Martin, Krizek, Nakayama, & Bradford, 1999 ). Another part of the project involved two of the team members’ integrated interpretive and critical methods to understand how whiteness is used as strategic rhetoric (Nakayama & Krizek, 1999 ). The volume included other scholars writing from different perspectives as well, and the editors attempted to bring together these various perspectives into a “coherent” picture about whiteness ideology. These scholars asked different questions and used different methods to investigate the same phenomena. Collectively, the research program told a richer and fuller story than any single study could have told. This example illustrates how different research methods can be used concurrently to advance understanding about intercultural phenomena.

Alternative Approaches to Studying Intercultural Communication

Intercultural research using the social scientific, interpretive, and critical methods have offered remarkable insights on a variety of intercultural phenomena. Each of these traditional approaches, however, uses a Euro-Western lens that is predominantly textocentric, privileging text, writing, and the lettered word in comparison to oral stories and visuals (Conquergood, 2002 ; Kim, 2002 ). We offer here two participatory approaches that, in some sense, hand over the power of the data to the participants. From these approaches, the ontology, epistemology, and axiology of the participants are more important than those of the researchers. Singhal, Harter, Chitnis, and Sharma ( 2007 ) explained that participation-based methodology allows for lateral communication between participants, creates a space for dialogue, focuses on the people’s needs, enables collective empowerment, and offers cultural-specific content. In contrast, they note that nonparticipatory methods allow top-down vertical communication, generally focus on individual behavior change, consider the donors’ and researchers’ goals of greater importance than community needs, and offer cultural-general information. This section discusses three participatory approaches: theater, photography, and community-based participatory research.

Participatory Theater

Based on the dialogic theorizing of Brazilian educator Paulo Freire ( 1970 ) and its application by Augusto Boal in his performative intervention, “Theater of the Oppressed” ( 1995 ), participatory theater can offer researchers an epistemology different from other research methods that rely on data from interviews and focus groups. This approach provides different kinds of data, discursive narratives that can be used to highlight some of the significant generative themes of the research participants.

The Theater of the Oppressed was developed in an effort to transform theater from the “monologue” of traditional performance into a “dialogue” between audience and stage. Boal ( 1995 ) experimented with many kinds of interactive theater. His explorations were based on the assumption that dialogue is the common, healthy dynamic between all humans and when a dialogue becomes a monologue, oppression ensues. Participatory theater is a research tool that produces generative and local knowledge, starting with the use of the body, the container of memory, emotions, and culture (Kaptani & Yuval-Davis, 2008 ). Theater has the ability to provide a useful connection to specific places as well as people. The encounter between the researcher and the researched in the theater space is outside the redundancy of everyday life. As a result, the researcher can see herself and her interactions between and with the researched in a way that is more distant than in everyday life, thus possibly making it easier to become reflexive.

Boal ( 1995 ) developed various forms of theater workshops and performances which aimed to meet the needs of all people for interaction, dialogue, critical thinking, action, and fun. For example, Forum Theater constitutes a series of workshops in which the participants are transformed from a passive audience into the double roles of actors and active audience. They construct dramatic scenes involving conflictual oppressive situations in small groups and show them to the other participants, who intervene by taking the place of the protagonists and suggesting better strategies for achieving their goals. One of the popular research tools used in Forum Theater is role-playing. Role-playing serves as a vehicle for analyzing power, stimulating public debate, and searching for solutions. Participants explore the complexity of the human condition and situate this knowledge in its cultural moment. The aim of the forum is not to find an ideal solution but to invent new ways of confronting problems. A second key tool is discussion. Following each intervention, audience members discuss the solution offered. A skilled facilitator encourages an in-depth discussion with the participants to generate ideas that will help to address issues under investigation.

Participatory Photography

Similarly, Paolo Friere is a pioneer in participatory photography. In 1973 , Freire and his team asked people living in a slum in Lima, Peru, to visualize “exploitation” by taking photographs (Singhal et al., 2007 ). One child took a photograph of a nail on a wall. While the photograph did not resonate with adults, many of the children strongly supported it. When asked to explain, it was learned that many of the boys in the neighborhood were shoeshine boys in the city. Since the shoeshine box was heavy and they could not carry it to the city, they rented a nail on the wall in one of the city shops. These shop owners charged the boys more than half of each day’s earnings as rent. The children expressed that the photo of the nail was the strongest symbol of exploitation. Friere and his team then used this photo to generate a discussion about exploitation and how the community members wished to address it.

Participatory photography, otherwise known as “photo voice” or “shooting back,” gives power to the participants, through photographs, to shape their own stories (Wang, 1999 ). Participatory photography has been used in a variety of contexts (slums, hospitals, schools, villages, etc.) and in different parts of the world (Singhal et al., 2007 ). For example, Briski and Kauffman ( 2004 ), in their Oscar-winning film, Born into Brothels: Calcutta’s Red Light Kids , taught the children of commercial sex workers how to take photographs. These children, then, took photos to depict their harsh reality. These powerful images became the foundation of this moving film. Another example is the work of Loignon et al. ( 2014 ) in Canada about the relation between impoverishment and lack of access to primary health care. The researchers recruited four family medicine residents and two medical supervisors to pursue their study. There were eight participants who came from economically underprivileged backgrounds trained in photographic techniques and photo voice philosophy. The researchers were able to realize the importance of primary health care professionals developing greater interpersonal and social acuity. They also realized that their patients were co-participants in the processes of diagnosis, prognosis, and medication. Finally, the researchers were also able to realize that they would be able to develop a greater competency by actually investing a part of their training time in the socioeconomic milieu of the patients they are to serve.

The implications of using participatory photography are significant. This method works best when the participants are given general directions and allowed to play with ideas. It is important for the participants to share their visual stories with the researchers. It is, however, critical for fellow participants in a community to share their stories with each other. The challenge of using photography is that it is, by nature, an intrusive process (Singhal et al., 2007 ). With terminology like “aim,” “shoot,” and “capture,” there can be a colonizing mentality in photography. It is particularly important that the participants be sensitive and reflective about how they take photographs of people and objects. While this may be difficult to accomplish across cultures, it is important to seek the permission of the participants before taking their photographs.

Community-Based Participatory Research

Community-based participatory research (CBPR) is a collaborative process where researchers and community members work together at all stages of the research process to address issues that are of importance to the community (Wallerstein et al., 2008 ). Rather than a top-down approach to health and social issues, CBPR focuses on a collaborative and bottom-up approach to identifying and defining problems and developing and implementing solutions (i.e., research “with” rather than “for” or “on”). CPBR is a preferred approach for researchers working with indigenous communities, other communities of color, or other communities facing disparities, which experience similar issues of mistrust for past research issues and social/health inequities. CBPR has goals of developing culturally centered research and interventions, building trust and synergy among partners, building the capacity of all members of the research team, changing power relations among communities and outside entities, developing sustainable change, and improving the social and health conditions of the community (Wallerstein et al., 2008 ).

CBPR is not a method, rather a philosophy of research. CBPR projects can include social scientific, interpretive, and critical approaches and often involve mixed methods. The specific methods meet the needs of the community and the research problem being addressed. The methods should follow key principles of CBPR, including: a) the project fits local/cultural beliefs, norms, and practices; b) the project emphasizes what is important in the community; c) the project builds on strengths in the community; d) the project balances research and social action; and e) the project disseminates findings to all partners and involves all partners in the dissemination process (Israel et al., 2008 ).

An example of CBPR is a project in Mysore, South India, addressing stigma and discrimination among men who have sex with men (MSM), many of whom were sex workers (Lorway et al., 2013 ). The project involved a collaboration of researchers and a sex workers collective in a long-term systematic process of knowledge production and action. The research approach involved training 10 community members as researchers who conducted interviews with MSM to understand their experiences. There were 70 in-depth interviews conducted in four days. Data analysis was completed with thematic analysis. The results provided a rallying point against stigma as the community cultivated its understanding of this concept and they mobilized to increase access to sexual health services.

The purpose of this article was to explore multidisciplinary methodological approaches to intercultural communication research. If our worldview shapes our reality, what we study and how we study phenomena is greatly influenced by our cultural frameworks. We described the traditional approaches to studying intercultural communication, namely, social scientific, interpretive, and critical perspectives. We identified the key ontological, epistemological, and axiological assumptions of each of these perspectives, offered an exemplar for each kind of perspective, types of data collected, and the methodological concerns in each framework. We then explained traditional ways to integrate the social scientific, interpretive, and critical perspectives, offered examples, and explicated the strengths of such integrations. We finally offered three alternative methodological approaches (participatory theater, participatory photography, and community-based participatory research) where the participants shape the scope of the study, interpret the meaning of the data, and offer practical implications for the study.


The early history of intercultural communication, including some discussion of research methods, has been covered well by Leeds-Hurwitz ( 1990 ) and Moon ( 1996 ). Leeds-Hurwitz reviewed the early foundation of intercultural communication, which can be traced to the work of Edward T. Hall in the Foreign Service Institute in the 1950s and 1960s. The focus in the earlier years was on descriptive linguistic analysis of micro communication practices (e.g., proxemics, kinesics, and verbal practices) of multiple cultures. These early roots of intercultural communication were influenced by anthropological study of culture (i.e., ethnography).

The 1970s saw the development of the field of intercultural communication, with a focus on culture as race, gender, nationality, and socioeconomic status (Moon, 1996 ). The research at this time also reflected the social issues of the 1970s. Methods of research were diverse but predominantly included social scientific and interpretive methods.

The late 1970s and the 1980s saw a change where the focus of culture became nationality and a large emphasis was placed on cross-cultural comparisons. There was a pursuit to develop and apply Western theories to non-Western contexts. Methodologically, the 1980s was dominated by social scientific approaches.

The 1990s brought some backlash against social scientific approaches from interpretive scholars. There was also a rise of critical scholarship which critiqued the social scientific research methods. A number of critical approaches were identified and were especially used to develop theoretical approaches for understanding intercultural communications.

The 2000s brought more balance and integration of the research approaches. The Journal of International and Intercultural Communication was founded in 2008 . The three editors of this journal to date (Tom Nakayama, Shiv Ganesh, and Rona Halualani) issued editorial statements about the scope of the journal respecting and including diverse methodological approaches.

Further Reading

  • Bhawuk, D. P. S. (2008). Science of culture and culture of science: Worldview and choice of conceptual models and methodology. The Social Engineer , 11 (2), 26–43.
  • Biggerstaff, D. L. , & Thompson, A. R. (2008). Interpretative phenomenological analysis: A qualitative methodology of choice in healthcare research. Qualitative Research in Psychology , 5 , 175–177.
  • Communication Methods & Measures. 2007–. Journal devoted to communication research methods; not exclusive to intercultural communication.
  • Fram, S. M. (2013). The constant comparative analysis method outside of grounded theory. The Qualitative Report , 18 (1), 4–8.
  • Israel, B. A. , Eng, E. , Schulz, A. J. , & Parker, E. (Eds.). (2013). Methods in community-based participatory research for health (2nd ed.). San Francisco: Jossey-Bass.
  • Lindlof, T. R. , & Taylor, B. C. (2011). Qualitative communication research methods (3rd ed.). Thousand Oaks, CA: SAGE.
  • Minkler, M. , & Wallerstein, N. (Eds.). (2008). Community-based participatory research for health: From process to outcomes . San Francisco: Jossey-Bass.
  • Otten, M. , & Geppert, J. (2010). Mapping the landscape of qualitative research on intercultural communication: A hitchhiker’s guide to the methodological galaxy. Forum: Qualitative Social Research , 10 (1), Article 52.
  • Tuhiwai-Smith, L. (1999). Decolonizing methodologies: Research and indigenous peoples . London: Zed Books.
  • Wagstaff, C. , Jeong, H. , Nolan, M. , Wilson, T. , Tweedlie, J. , Phillips, E. , & Holland, F. (2014). The accordian and the deep bowl of spaghetti: Eight researchers’ experiences of using IPA as a methodology. The Qualitative Report , 19 (47), 1–4.
  • Asante, M. K. , Miike, Y. , & Jing, Y. (2008). Introduction. Issues and challenges in intercultural communcation scholarship. In M. K. Asante , Y. Miike , & Y. Jing (Eds.), The global intercultural communication reader (pp. 1–8). New York: Routledge.
  • Baig, N. , Ting-Toomey, S. , & Dorjee, T. (2014). Intergenerational narratives on face: A South Asian Indian American perspective. Journal of International and Intercultural Communication , 7 , 127–147.
  • Boal, A. (1995). Theater of the oppressed . New York: Theater Communications Group.
  • Braun, V. , & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology , 3 , 77–101.
  • Brinson, M. E. , & Stohl, M. (2012). Media framing of terrorism: Implications for public opinion, civil liberties, and counterterrorism policies. Journal of International and Intercultural Communication , 5 , 270–290.
  • Briski, Z. (Prod. & Dir.), & Kauffman, R. (Prod. & Dir.) (2004). Born into Brothels: Calcutta’s Red Light Kids [Motion Picture]. India: THINKFilm
  • Cai, D. A. , Wilson, S. R. , & Drake, L. E. (2000). Culture in the context of intercultural Negotiation: Individualism-collectivism and paths to integrative agreements. Human Communication Research , 26 , 591–617.
  • Carbaugh, D. (2007). Cultural discourse analysis: Five communication practices and intercultural encounters. Journal of Intercultural Communication Research , 36 (3), 67–182.
  • Cissel, M. (2014). Media framing: A comparative content analysis on mainstream and alternative news coverage of Occupy Wall Street. The Elon Journal of Undergraduate Research in Communications , 3 (1), 68–72.
  • Chen, Y-W. , Simmons, N. , & Kang, D. (2015). My family isn’t racist—however …”: Multiracial/multicultural Obama-ism as an ideological barrier to teaching intercultural communication. Journal of International and Intercultural Communication , 8 , 167–186.
  • Cohen, L. , Manion, L. , & Morrison, K. (2007). Research methods in education (6th ed.). London: Routledge.
  • Collier, M. J. (2005). Theorizing cultural identifications: Critical updates and continuing Evolution. In W. B. Gudykunst (Ed.), Theorizing about intercultural communication (pp. 235–256). Thousand Oaks, CA: SAGE.
  • Collins, P. H. (1990). Black feminist thought: Knowledge, consciousness, and the politics of empowerment . Boston: Unwin Hyman.
  • Conquergood, L. (2002). Lethal theatre: Performance, punishment, and the death penalty . Theatre Journal , 54 (3), 339–367.
  • Cook, J. A. , & Fonow, M. M. (1984). Am I my sister’s gatekeeper? Cautionary tales from the academic hierarchy. Humanity and Society , 8 , 442–452.
  • Corbin, J. , & Strauss, A. (2008). Basics of qualitative research: Techniques to developing grounded theory (3rd ed.). Thousand Oaks, CA: SAGE.
  • Cresswell, J. W. , & Plano Clark, V. L. (2011). Designing and conducting mixed methods research (2nd ed.). Thousand Oaks, CA: SAGE.
  • Denzin, N. K. (2001). The reflexive interview and a performative social science. Qualitative Research, 1(1) , 28–34.
  • Dixon, T. L. , & Azocar, C. L. (2006). The representation of juvenile offenders by race on Los Angeles area television news. The Howard Journal of Communications , 17 , 143–161.
  • DuBois, W. E. B. (1969). The souls of black folks. New York: New American Library.
  • Fairclough, N. (1995). Critical discourse analysis: The critical study of language. London: Longman.
  • Fletcher, C. V. , Nakazawa, M. , Chen, Y. , Oetzel, J. G. , Ting-Toomey, S. , Chang , & Zhang, Q. (2014). Establishing cross-cultural measurement equivalence of scales associated with face-negotiation theory: A critical issue in cross-cultural comparisons. Journal of International and Intercultural Communication , 7 , 148–169.
  • Freire, P. (1970). Pedagogy of the oppressed . New York: Continuum.
  • Golafshani, N. (2003). Understanding reliability and validity in qualitative research. The Qualitative Report , 8 , 597–607.
  • Gudykunst, W. B. (2003a). Forward. In W. B. Gudykunst (Ed.), Cross-cultural and Intercultural Communication (pp. vii–ix). Thousand Oaks, CA: SAGE.
  • Gudykunst, W. B. (2003b). Issues in cross-cultural research. In W. B. Gudykunst (Ed.), Cross-cultural and intercultural communication . Thousand Oaks, CA: SAGE.
  • Hall, J. K. (2002). Teaching and researching language and culture . London: Pearson Education.
  • Han, B. , & Cai, D. A. (2010). Face goals in apology: A cross-cultural comparison of Chinese and U.S. Americans. Journal of Asian Pacific Communication , 20 , 101–123.
  • Hanasono, L. K. , Chen, L. , & Wilson, S. R. (2014). Identifying communities in need: Examining the impact of acculturation on perceived discrimination, social support, and coping amongst racial minority members in the United States. Journal of International and Intercultural Communication , 7 , 216–237.
  • Harding, S. (1998). Is science multicultural? Postcolonialsim, feminism, and epistemologies . Bloomington: Indiana University Press.
  • Hermans, H. J. M. , & Kempen, H. J. G. (1998). Moving cultures: The perilous problem of cultural dichotomies in a globalizing society. American Psychologist , 53 , 1111–1120.
  • Hurtado, A. (1996). The color of privilege: Three blasphemies on race and feminism . Ann Arbor: University of Michigan Press.
  • Irani, L. , Vertesi, J. , Dourish, P. , Philip, K. , & Grinter, R. E. (2010). Postcolonial computing: A lens on design and development. Proceedings of the ISGCHI conference on human factors in computing systems. CHI’10, 1311–1320.
  • Israel, B. A. , Schulz, A. J. , Parker, E. A. , Becker, A. B. , & Guzman, J. R. (2008). Critical issues in developing and following CBPR principles. In M. Minkler & N. Wallerstein (Eds.), Community-based participatory research for health: From process to outcomes (2nd ed., pp. 47–66). San Francisco: Jossey-Bass.
  • Kaplan, Abraham (1964). The conduct of inquiry: Methodology for behavioral science . San Francisco: Chandler.
  • Kaptani, E. , & Yuval-Davis, N. (2008). Participatory theatre as a research methodology . Sociological Research On Line , 13 , 2.
  • Kim, M. S. (2002). Multicultural perspectives on human communication: Implications for theory and practice . Thousand Oaks, CA: SAGE.
  • Kelleher, W. J. (2013). Formal axiology and the philosophy of social science; esp., political science . Available at SSRN: or
  • Klein, H. , & Shiffman, K. (2006). Race-related content of animated cartoons. The Howard Journal of Communications , 17 , 163–182.
  • Koopman, S. (2011). Alter-geopolitics: Other securities are happening. Geoforum , 3 , 274–284.
  • Lazar, M. M. (2005). Politicizing gender in discourse: Feminist critical discourse analysis as political perspective and praxis. In M. M. Lazar (Ed.), Feminist critical discourse analysis: Gender, power and ideology in discourse (pp. 1–27). New York: Palgrave Macmillan.
  • Leeds-Hurwitz, W. (1990). Notes in the history of intercultural communication: The Foreign Service Institute and the mandate for intercultural training. Quarterly Journal of Speech , 76 , 262–281.
  • Levine, T. R. , Park, H. S. , & Kim, R. K. (2007). Some conceptual and theoretical challenges for cross-cultural communication research in the 21st century. Journal of Intercultural Communication Research , 36 , 205–221.
  • Lincoln, Y. S. , & Guba, E. G. (1985). Naturalistic inquiry . Beverly Hills, CA: SAGE.
  • Loignon, C. , Boudreault-Fournier, A. , Truchon, K. , Labrousse, Y. , & Fortin, B. (2014). Medical residents reflect on their prejudices toward poverty: A photovoice training project .
  • Lorway, R. , Thompson, L. H. , Lazarus, L. , du Plessis, E. , Pasha, P. , Mary, P. F. , et al. (2013). Going beyond the clinic: Confronting stigma and discrimination among men who have sex with men in Mysore through community-based participatory research. Critical Public Health , 24 (1), 1–15.
  • Martin, J. N. , Krizek, R. L. , Nakayama, T. K. , & Bradford, L. (1999). What do white people want to be called? A study of self-labels for white Americans. In T. K. Nakayama & J. N. Martin (Eds.), Whiteness: The communication of social identity (pp. 27–50). Thousand Oaks, CA: SAGE.
  • Mohanty, C. T. (1991). Cartographies of struggle: Third world women and the politics of feminism. In C. T. Mohanty , A. Russo , & L. Torres (Eds.), Third world women and the politics of feminism (pp. 1–50). Bloomington: Indiana University Press.
  • Moon, D. G. (1996). Concepts of culture. Implications for intercultural communication research. Communication Quarterly , 44 , 70–84.
  • Nakayama, T. K. , & Krizek, R. L. (1999). Whiteness as strategic rhetoric. In T. K. Nakayama & J. N. Martin (Eds.), Whiteness: The communication of social identity (pp. 87–106). Thousand Oaks, CA: SAGE.
  • Nakayama, T. K. , & Martin, J. N. (Eds.). (1999). Whiteness: The communication of social identity . Thousand Oaks, CA: SAGE.
  • O’Connor, M. K. , Netting, F. E. , & Thomas, M. L. (2008). Grounded theory: Managing the challenge for those facing institutional review board oversight. Qualitative Inquiry , 14 (1), 11–15.
  • Oetzel, J. G. (1998). Culturally homogeneous and heterogeneous groups: Explaining communication processes through individualism-collectivism and self-construal. International Journal of Intercultural Relations , 22 , 135–161.
  • Oetzel, J. G. (1998). Explaining individual communication processes in homogeneous and heterogeneous groups through individualism-collectivism and self-construal. Human Communication Research , 25 , 202–224.
  • Oetzel, J. G. , & Ting-Toomey, S. (2003). Face concerns in interpersonal conflict: A cross-cultural empirical test of the face-negotiation theory. Communication Research , 30 , 599–624.
  • Pathak, A. (2010). Opening my voice, claiming my space: Theorizing the possibilities of postcolonial approaches to autoethnography. Journal of Research Practice , 6 (1), 4–11.
  • Perko, T. , Turcanu, C. , Geenen, D. , Mamani, N. , & Rooy, Van L. (2011). Media content analysis of the Fukushima accident in two Belgian newspapers . Report: Belgian Nuclear Research Centre. Available:
  • Philipsen, G. (1975). Speaking “like a man” in Teamsterville: Culture patterns of role enactment in an urban neighborhood. Quarterly Journal of Speech , 61 , 13–22.
  • Radford, M. L. , Radford, P. G. , Connaway L. S. , & DeAngelis, J. A. (2011). On virtual face-work: An ethnography of communication approach to a live chat. Library Quarterly , 81 , 431–453.
  • Rao, N. , Singhal, A. , Ren, L. , & Zhang, J. (2001). Is the Chinese self-construal in transition? Asian Journal of Communication , 11 , 68–95.
  • Schmid, A. P. (2013). Radicalisation, de-radicalisation, counter-radicalisation: A conceptual discussion and literature review . International Centre for Counter-Terrorism. Available:
  • Scotland, J. (2012). Exploring the philosophical underpinnings of research: Relating ontology and epistemology to the methodology and methods of the scientific, interpretive, and critical research paradigms. English Language Teaching , 5 (9), 9–16.
  • Singhal, A. , Harter, L. , Chitnis, K. , & Sharma, D. (2007). Participatory photography as theory, method, and praxis: Analyzing an entertainment-education project in India. Critical Arts , 21 , 212–227.
  • Spivak, G. C. (1999). A critique of postcolonial reason: Toward a history of the vanishing present . Cambridge, MA: Harvard University Press.
  • Sponcil, M. , & Gitimu, P. (2010). Use of social media by college students: Relationship to communication and self-concept. Journal of Technological Research , 4 , 1–13.
  • Straub, J. , & Weidemann, D. (2006). Psychology, culture, and the pursuit of meaning: An introduction. In J. Straub , D. Weidemann , C. Kölbl , & B. Zielke (Eds.), Pursuit of meaning. Advances in cultural and cross-cultural psychology (pp. 11–20). Bielefeld: Transcript Verlag.
  • Szabo, E. (2007). Participative management and culture. A qualitative and integrative study in five European countries . Frankfurt: Peter Lang.
  • Van Dijk, T. (2008). Discourse in context: A sociocognitive approach. Cambridge, U.K.: Cambridge University Press.
  • Wallerstein, N. , Oetzel, J. , Duran, B. , Belone, L. , Tafoya, G. , & Rae, R. (2008). CBPR: What predicts outcomes? In M. Minkler & N. Wallerstein (Eds.), Community-based participary research for health: From process to outcomes (2nd ed., pp. 371–392). San Francisco: Jossey-Bass.
  • Wang, C. (1999). Photo voice: A participatory action research strategy applied to women’s health. Journal of Women’s Health , 8 , 185–192.
  • Willink, K. , Gutierrez-Perez, R. , Shukri, S. , & Stein, L. (2014). Navigating with the stars: Critical qualitative methodological constellations for critical intercultural communication research. Journal of International and Intercultural Communication , 7 , 289–316.
  • Zhang, Q. , & Oetzel, J. G. (2006). Constructing and validating a teacher immediacy scale from a Chinese perspective. Communication Education , 55 , 218–241.
  • Zhang, Q. , Oetzel, J. G. , Gao, X. , Wilcox, R. , & Takai, J. (2007). Teacher immediacy scales: Testing for validity across cultures. Communication Education , 56 , 228–248.

Related Articles

  • Intercultural Competence
  • Intercultural Friendships
  • Globalizing and Changing Culture
  • Bella Figura: Understanding Italian Communication in Local and Transatlantic Contexts

Printed from Oxford Research Encyclopedias, Communication. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 04 June 2024

  • Cookie Policy
  • Privacy Policy
  • Legal Notice
  • Accessibility
  • [|]

Character limit 500 /500

research methods for communication

  • Register or Log In
  • 0) { document.location='/search/'+document.getElementById('quicksearch').value.trim().toLowerCase(); }">

Wrench et al, Quantitative Research Methods for Communication: A Hands-On Approach 4e

  • Communication/Media Studies/Journalism

research methods for communication

Quantitative Research Methods for Communication 4e Instructor Resources

Wrench et al

Instructor Resources for Wrench et al, Quantitative Research Methods for Communication 4e

Quantitative Research Methods for Communication 4e Student Resources

Student Resources for Wrench et al, Quantitative Research Methods for Communication 4e

Select your Country

  • Technical Support
  • Find My Rep

You are here

The Sage Handbook of Qualitative Research in Organizational Communication

The Sage Handbook of Qualitative Research in Organizational Communication

  • Boris H. J. M. Brummans - Université de Montréal, Canada
  • Bryan C. Taylor - University of Colorado Boulder, USA
  • Anu Sivunen - University of Jyväskylä, Finland
  • Description

The Sage Handbook of Qualitative Research in Organizational Communication  is a state-of-the-art resource for scholars, students, and practitioners seeking to deepen their understanding and expertise in this dynamic field. Written by a global team of established and emerging experts, this Handbook provides a comprehensive exploration of the field’s foundational traditions of epistemology and theory, as well as its latest methodologies, methods, issues, and debates.

The volume reflects a diverse range of approaches (e.g., mixed-methods, ethnographic, rhetorical, pragmatist, phenomenological, feminist, critical race, postcolonial, queer, and engaged), and covers a broad spectrum of topics ranging from data collection and analysis, to representation. Additionally, this Handbook addresses emerging trends such as digital forensics, post-qualitative research, and the transformative impact of COVID-19 on the conduct of qualitative research in organizational communication. As the first volume of its kind in this field, The Sage Handbook of Qualitative Research in Organizational Communication is a cornerstone text for scholars, students, and practitioners interested in understanding the vital role of communication in organizational life. Part 1: Approaches to Qualitative Organizational Communication Research  Part 2: Data Collection in Qualitative Organizational Communication Research: Methods and Issues Part 3: Data Analysis and Representation in Qualitative Organizational Communication Research: Methods and Issues Part 4: The Future of Qualitative Organizational Communication Research

See what’s new to this edition by selecting the Features tab on this page. Should you need additional information or have questions regarding the HEOA information provided for this title, including what is new to this edition, please email [email protected] . Please include your name, contact information, and the name of the title for which you would like more information. For information on the HEOA, please go to .

For assistance with your order: Please email us at [email protected] or connect with your SAGE representative.

SAGE 2455 Teller Road Thousand Oaks, CA 91320

The Sage Handbook of Qualitative Research in Organizational Communication is an essential resource for management scholars, practitioners, and students alike. This groundbreaking handbook is a comprehensive guide that navigates the vast landscape of qualitative research within the realm of organizational communication. With contributions from leading experts in the field, it provides invaluable insights, methodologies, and theoretical frameworks helping to understand the complex dynamics of communication in organizational settings. It offers a diverse range of perspectives and approaches, inviting readers to explore the intricacies of organizational communication through a qualitative lens. This handbook is an indispensable companion for anyone seeking to delve deeper into this fascinating field.

Communication scholars have developed many rich and rigorous qualitative methods for capturing, analyzing, and interpreting the communicative aspects of organizing. It is exciting to see these methods brought together in a comprehensive handbook that features both classical approaches and reflections on new forms of data and analysis. Boris Brummans, Bryan Taylor, and Anu Sivunen have put together a truly wonderful resource that I will return to again and again.

This handbook is an authoritative survey of qualitative approaches to organizational communication that summarizes the state of the art and advances new insights and ideas. Its authors span the panoply of takes on qualitative research on organizational communication and are literally an all-star cast. Experienced academics and graduate students alike will benefit from its many perspectives and inquiries into key issues.

This comprehensive handbook has all the ingredients to become the main go-to resource for qualitative research in organizational communication studies. I highly recommend it to anyone who aims to broaden their horizon and expand their methodological tool kit, whether scholars, students, or practitioners.

Preview this book

Select a purchasing option.

SAGE Knowledge Promotion

This title is also available on SAGE Knowledge , the ultimate social sciences online library. If your library doesn’t have access, ask your librarian to start a trial .

research methods for communication

  • Research Measures

Dr. Jason S. Wrench

  • Journal Articles
  • Conferences And Talks
  • Book Chapters

Kendall-Hunt Books

Dr. Wrench has had the opportunity to work on 5 edited case study collections and two textbooks with Kendall-Hunt.

SUNY New Paltz

Dr. Wrench is an Associate Professor and former chair of the Department of Communication at the State University of New York at New Paltz.

Dr. Wrench’s Research

Dr. Wrench has published over 35 peer-reviewed research articles. You can find full text copies of all but his most recent publications here on his website. To ones that do not have full-text, links to the publishers' websites are provided.

Directory of Communication Measures

The Directory of Communication Mental Measures

In 2010 the National Communication Association published The Directory of Communication Related Mental Measures: A Comprehensive Index of Research Scales, Questionnaires, Indices, Measures, and Instruments by Jason Wrench, Doreen Jowi, and Alan Goodboy. Although this book is no longer in print, the content is now available as a searchable Wiki. Furthermore, new measures are always being added.

Quantitative Research Methods For Communication: A Hands-On Approach

Quantitative Research Methods for Communication: A Hands-On Approach

Quantitative Research Methods for Communication: A Hands-On Approach by Jason Wrench, Candice Thomas-Maddox, Virginia Richmond, and James McCroskey is currently in the 3rd edition and the 4th edition is arriving in 2019. The textbook's website contains a number of helpful video tutorials and links that help new researchers explore the world of quantitative communication research.

Training and Development: The Intersection of Communication and Talent Development in the Modern Workplace

Training and Development: The Intersection of Communication and Talent Development in the Modern Workplace The second edition will be available in Spring 2023.

This comprehensive book on the field of Training and Development was authored by Jason Wrench, Danette Johnson, and Maryalice Citera. The book walks learners through the process of creating both face-to-face and eLearning programs. Other features include information on human performance improvement, training management, and project management. The corresponding website has a number of online tools discussed in the textbook itself to provide learners a full toolkit to get them started in the world of Talent Development

Current Research

Dr. Jason S. Wrench is always conducting research related to his interests in intercultural, interpersonal, instructional, and organizational communication. To learn about some of his past and current projects click here:

Jason Wrench

Introduction to Dr. Jason S. Wrench

Jason S. Wrench, Ed.D. is a Professor at SUNY New Paltz.

About Current Publications

As a researcher, I am constantly engaged in a wide range of different research topics related to areas related to communication and technology, communication education, communication and religion, instructional communication, organizational communication, and quantitative research methods. 

Currently, my primary research partner, Dr. Narissra M. Punyanunt-Carter, have been working on a number of projects related communication and religion and communication and technology. We’ve been lucky to know each other for over 20 years and publish both research articles and books during that time.

Investigating the relationships among college students’ satisfaction, addiction, needs, communication apprehension, motives, and uses & gratifications with Snapchat

“i am spiritual, not religious”: examination of the religious receiver apprehension scale, affective learning: evolving from values and planned behaviors to internalization and pervasive behavioral change.

JASON S. WRENCH (Ed.D., West Virginia University) is an associate professor and chair of the Department of Communication at the State University of New York at New Paltz. Dr. Wrench specializes in workplace learning and performance, or the intersection of i

EdD in Curriculum & Instruction and Communication Studies

Ma in communication studies.

December 1997

  • What is New
  • Download Your Software
  • Behavioral Research
  • Software for Consumer Research
  • Software for Human Factors R&D
  • Request Live Demo
  • Contact Sales

Sensor Hardware

Man wearing VR headset

We carry a range of biosensors from the top hardware producers. All compatible with iMotions

iMotions for Higher Education

Imotions for business.

research methods for communication

How Does Environmental Ergonomics Affect Behavior?

Ergonomics Testing

Morten Pedersen

research methods for communication

Can You Build Your Own Eye Tracking Glasses?

News & events.

  • iMotions Lab
  • iMotions Online
  • Eye Tracking
  • Eye Tracking Screen Based
  • Eye Tracking VR
  • Eye Tracking Glasses
  • Eye Tracking Webcam
  • FEA (Facial Expression Analysis)
  • Voice Analysis
  • EDA/GSR (Electrodermal Activity)
  • EEG (Electroencephalography)
  • ECG (Electrocardiography)
  • EMG (Electromyography)
  • Respiration
  • iMotions Lab: New features
  • iMotions Lab: Developers
  • EEG sensors
  • Sensory and Perceptual
  • Consumer Inights
  • Human Factors R&D
  • Work Environments, Training and Safety
  • Customer Stories
  • Published Research Papers
  • Document Library
  • Customer Support Program
  • Help Center
  • Release Notes
  • Contact Support
  • Partnerships
  • Mission Statement
  • Ownership and Structure
  • Executive Management
  • Job Opportunities


  • Newsletter Sign Up

What is Communication Research?

Bryn Farnsworth

Bryn Farnsworth

Table of Contents

Let’s start things off with a hopefully rather non-controversial assumption: you’re reading this text. Now, don’t worry, the rest of the post won’t continue to list such inanely clear truths, but this statement serves to highlight what’s occurring right now – communication.

It’s a rather one-sided process currently, with me doing the writing and you doing the reading, but the process of sending a message and that message being understood is occurring. There are endless ways of looking further at this – would you still be reading if you were on a different website? What about if you’re tired, or angry? How would whatever else you’ve read today – or ever – affect how you react to this information? This in essence is what communication research is about – how messages are sent, and how they are received.

At its broadest, communication research is concerned with identifying, exploring, and measuring the factors that surround communication, in any form and regarding any topic . Often from a theory-driven perspective, but increasingly with empirically-grounded methods. Want to know how to make political messaging more effective? Increase the appeal of advertising ? Make people adhere to a health campaign? Communication research these answers.

Below, we will discuss and define communication research further, the research that has shaped the field, and where the field is going.

Definition of Communication Research

As a field of study, communication research dates back either 2000 years or 100 years, depending on your level of pedantry. The study of rhetoric was a hot topic in ancient Greece, and shares some commonalities with the modern form, yet clearly much has changed. The field now focuses on gathering empirical data, and builds theories that help understand the complexity of communication on many levels. In a sense it has less interest in the linguistic style of debating philosophers, and more interest in the groups of people that might be listening.

Socrates death painting

History of Communication Research

One of the most influential books that helped give rise to modern communication research was “ Social Organization: a Study of the Larger Mind ” by Charles Cooley, published in 1909 [1]. Described by one reviewer as “a series of essays on fundamental sociological problems, written in delightful literary style, and with keen and sound psychological insight” and that “Professor Cooley gives, for the first time in sociological literature, strange as it may seem, full and adequate recognition of ‘communication’ as a fundamental fact in the social life” [2].

This book, with a delightful literary style , would set the stage for the work of other academics with an interest in communication, and ultimately the creation of the first academic departments with a clear focus on the field.

In 1952, Bernard Berelson released “ Content Analysis in Communication Research ” – a book that proved pivotal not only to communication researchers of the time, but also had a broader impact [3, 4]. Written in a way that was – according to one reviewer at the time – “unusually lucid for a social science publication”, the book describes the ways in which media and communication are compared, and explores the methods that are used to carry out those comparisons. The book ultimately helped shift the field towards a more quantitative, scientific approach.

In the 1960s and 1970s, social unrest brought about social change, and communication researchers looked more closely at the surrounding language. They explored the systems of thought and discourse that had traditionally been in place, how they were changing, and what that might mean for the future of communication [5]. This occurred alongside the continual expansion of mass communication methods – TV and radio continued their dominance of message-spreading in the western world.

The shift into empirical methodology continued. While theoretical discussions of communication remained (and remain) central to the field, the introduction of data-driven, quantified assessments became an increasingly routine aspect of communication research. The book “ Mass Communication Research Methods ”, released in 1998, helped cement this as standard, defining the experimental methods of the day [6].

These research methods – focus groups , observations, and surveys – have now long been central to the field, yet the next step in empirical quantification is already emerging. Continuing with the steps towards quantification and more thoroughly empirical approaches, new unbiased tools are now being used as a way to incisively measure the processes surrounding communication, to test theories, and to advance understanding further. But what does this look like?

New Methods for Communication Research

Eye tracking has become one of the most widely used technologies within communication research, largely as it “gives communication scholars the opportunity to examine more precisely how much visual attention has been paid to information” [7].

participant using the eye tracking webcam in front of the screen

In 2016, researchers from the University of Amsterdam carried out the first retrospective study examining the use of eye tracking technology within communication research [7]. They found that the majority of studies within communication using eye tracking had focused on advertising research , yet public health, language, and computer-mediated communication were also areas that had been looked at. They also conclude that “ that eye tracking has much more potential in communication research”.

research methods for communication

One example of this potential being seized upon is found in research by researchers from Ohio State University and the University of Illinois Urbana–Champaign, who developed eye tracking metrics to assess automatic stereotyping [8]. By using a gaze-contingency task , they were able to show that stereotype-congruent fixations were decreased for those with a higher political knowledge score.

Political Communication Research

The research showed that participants who are categorized as knowledgeable about politics were more capable of “moderating automatic responses” – adding a new layer of understanding to how political communication can impact reported and actual responses. The researchers go on to state that this “implies that the influence of automatic processes on political thinking is conditional” – meaning that our response to political communication may be less automatic than previously thought.

Responsive Media Messages

While these studies have used eye tracking to measure attention, other communication research has used a combination of methods. Researchers from Texas Tech University used facial electromyography (fEMG), electrocardiography (ECG) , and electrodermal activity (EDA) in order to assess affect in response to media messages [9].

They found that fEMG data provided reliable data regarding emotional state, while heart rate data collected from ECG indicated that negative messages received more attention than positive messages. The skin conductance data collected from EDA provided data that, together with a memory test, showed that the arousal level experienced was a greater predictor of memory retention for the media exposure, as compared to the valence experienced.

Increasing Engagement

Other research has also used arousal in order to understand the response to communication (for a review of some of these studies, see [10]). For example, researchers from Indiana University and the University of Wisconsin-Madison investigated responses to the number of edits within media using EDA and ECG [11]. They find that an increase of edits within the media can increase the encoding of the message without causing too much cognitive load , suggesting that media should feature a larger number of edits (where appropriate) to increase engagement.

Free 52-page Human Behavior Guide

For Beginners and Intermediates

  • Get accessible and comprehensive walkthrough
  • Valuable human behavior research insight
  • Learn how to take your research to the next level

research methods for communication

[1] Cooley, C. H. (1962). Social Organization: A Study of the Larger Mind . New York: Schocken (first published 1909).

[2] Ellwood, C. A. (1910). Social Organization: A Study of the Larger Mind. Charles Horton Cooley. The International Journal of Ethics , 20 : 2 ,  228-230.

[3] Berelson, B. (1952). Content analysis in communication research . Glencoe, IL: Free Press.

[4] Bauer, M. (2000) “Classical Content Analysis: A Review,” in M. Bauer and G. Gaskell (eds.), Qualitative Researching with Text, Image and Sound — A Handbook . London: SAGE. pp. 131—150.

[5] Park, D. W., & Pooley, J. (2008). The history of media and communication research: Contested memories . New York: Peter Lang.

[6] Hansen, A., Cottle, S., Negrine, R. and Newbold, C. (1998). Mass Communication Research Methods . London: Macmillan.

[7] Bol, N., Boerman, S. C., Romano Bergstrom, J. C., & Kruikemeier, S. (2016). An overview of how eye tracking is used in communication research. In M. Antona & C. Stephanidis (Eds.), International conference on universal access in human-computer interaction . Proceedings HCII 2016, Part I, LNCS 9737 ed. (pp. 421–429). Switzerland: Springer International Publishing.

[8] Coronel, J. C., & Federmeier, K. D. (2016). The Effects of Gender Cues and Political Sophistication on Candidate Evaluation: A Comparison of Self-Report and Eye Movement Measures of Stereotyping. Communication Research , 43(7), 922-944. doi:10.1177/0093650215604024.

[9] Bolls, P.D., Lang, A., & Potter, R.F. (2001). The effects of message valence and listener arousal on attention, memory, and facial muscular responses to radio advertisements. Communication Research , 28, 627-651.

[10] Ravaja, N. (2004). Contributions of psychophysiology to media research: Review and recommendations. Media Psychology , 6, 193-235.

[11] Lang, A., Zhou, S., Schwartz, N., Bolls, P. D., & Potter, R. F. (2000). The effects of edits on arousal, attention, and memory for television messages: When an edit is an edit can an edit be too much? Journal of Broadcasting & Electronic Media , 44(1), 94-109.

Last edited

About the author

See what is next in human behavior research

Follow our newsletter to get the latest insights and events send to your inbox.

Related Posts

Illustration of social behavior showing an abstract crowd of people

What is Social Behavior?

research methods for communication

What is Covert Behavior?

research methods for communication

What is Overt Behavior?

research methods for communication

Beyond Words: Behavioral Expressions in Communication

You might also like these.

research methods for communication

Exploring Mobile Eye Trackers: How Eye Tracking Glasses Work and Their Applications

research methods for communication

Understanding Screen-Based Eye Trackers: How They Work and Their Applications

Product guides.

Human Factors and UX

Case Stories

Explore Blog Categories

Best Practice

Collaboration, product news, research fundamentals, research insights, 🍪 use of cookies.

We are committed to protecting your privacy and only use cookies to improve the user experience.

Chose which third-party services that you will allow to drop cookies. You can always change your cookie settings via the Cookie Settings link in the footer of the website. For more information read our Privacy Policy.

  • gtag This tag is from Google and is used to associate user actions with Google Ad campaigns to measure their effectiveness. Enabling this will load the gtag and allow for the website to share information with Google. This service is essential and can not be disabled.
  • Livechat Livechat provides you with direct access to the experts in our office. The service tracks visitors to the website but does not store any information unless consent is given. This service is essential and can not be disabled.
  • Pardot Collects information such as the IP address, browser type, and referring URL. This information is used to create reports on website traffic and track the effectiveness of marketing campaigns.
  • Third-party iFrames Allows you to see thirdparty iFrames.

research methods for communication

  • Words, Language & Grammar

Kindle app logo image

Download the free Kindle app and start reading Kindle books instantly on your smartphone, tablet, or computer - no Kindle device required .

Read instantly on your browser with Kindle for Web.

Using your mobile phone camera - scan the code below and download the Kindle app.

QR code to download the Kindle App

Image Unavailable

Quantitative Research Methods for Communication: A Hands-On Approach

  • To view this video download Flash Player

Follow the author

Jason S. Wrench

Quantitative Research Methods for Communication: A Hands-On Approach 4th Edition

Purchase options and add-ons.

  • ISBN-10 0190861061
  • ISBN-13 978-0190861063
  • Edition 4th
  • Publisher Oxford University Press
  • Publication date November 15, 2018
  • Language English
  • Dimensions 9.9 x 0.9 x 8 inches
  • Print length 672 pages
  • See all details

Amazon First Reads | Editors' picks at exclusive prices

Frequently bought together

Quantitative Research Methods for Communication: A Hands-On Approach

Customers who bought this item also bought

Publication Manual (OFFICIAL) 7th Edition of the American Psychological Association

Editorial Reviews

Book description, about the author, product details.

  • Publisher ‏ : ‎ Oxford University Press; 4th edition (November 15, 2018)
  • Language ‏ : ‎ English
  • Paperback ‏ : ‎ 672 pages
  • ISBN-10 ‏ : ‎ 0190861061
  • ISBN-13 ‏ : ‎ 978-0190861063
  • Item Weight ‏ : ‎ 2.6 pounds
  • Dimensions ‏ : ‎ 9.9 x 0.9 x 8 inches
  • #494 in Communication Reference (Books)
  • #1,499 in Communication & Media Studies
  • #17,041 in Unknown

About the author

Jason s. wrench.

Jason Wrench writes. I write a lot. I’m the author of 12 Days of Murder (November 2021) and Till Death Do Us Wed (February 2022), and the Up on the Farm trilogy: Finding a Farmer (August, 2022), Bewitched by the Barista (September, 2022), and Sanctuary for the Surgeon (January, 2023), and Wolf Island (October, 2022) all with Pride Publishing. I’m also the author of The Veil and Jekyll/Hyde. And I’m the man behind the cozy mystery penname, J. J. Justice.

When I’m not writing novels, I’m a college professor at SUNY New Paltz in the Department of Communication. In that capacity, I’ve authored or edited twenty academic books, thirty-five plus research articles, and numerous chapters in other books.

In my downtime, I love reading/writing, Broadway, coffee, and my puggle, Max (6-year-old) and Teddy (16-year-old).

I’m a member of the Romance Writers of America and the Textbook and Academic Author Association.

Customer reviews

Customer Reviews, including Product Star Ratings help customers to learn more about the product and decide whether it is the right product for them.

To calculate the overall star rating and percentage breakdown by star, we don’t use a simple average. Instead, our system considers things like how recent a review is and if the reviewer bought the item on Amazon. It also analyzed reviews to verify trustworthiness.

  • Sort reviews by Top reviews Most recent Top reviews

Top reviews from the United States

There was a problem filtering reviews right now. please try again later..

research methods for communication

Top reviews from other countries

research methods for communication

  • Amazon Newsletter
  • About Amazon
  • Accessibility
  • Sustainability
  • Press Center
  • Investor Relations
  • Amazon Devices
  • Amazon Science
  • Sell on Amazon
  • Sell apps on Amazon
  • Supply to Amazon
  • Protect & Build Your Brand
  • Become an Affiliate
  • Become a Delivery Driver
  • Start a Package Delivery Business
  • Advertise Your Products
  • Self-Publish with Us
  • Become an Amazon Hub Partner
  • › See More Ways to Make Money
  • Amazon Visa
  • Amazon Store Card
  • Amazon Secured Card
  • Amazon Business Card
  • Shop with Points
  • Credit Card Marketplace
  • Reload Your Balance
  • Amazon Currency Converter
  • Your Account
  • Your Orders
  • Shipping Rates & Policies
  • Amazon Prime
  • Returns & Replacements
  • Manage Your Content and Devices
  • Recalls and Product Safety Alerts
  • Conditions of Use
  • Privacy Notice
  • Consumer Health Data Privacy Disclosure
  • Your Ads Privacy Choices

City University of Hong Kong

  • examine basic concepts of communication research;
  • introduce specific data collection methods such as survey, content analysis, experimental design and qualitative methods;
  • practice statistical techniques including sampling, descriptive and inferential analyses;
  • oral and written presentation of research results. 

City University of Hong Kong

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • View all journals
  • My Account Login
  • Explore content
  • About the journal
  • Publish with us
  • Sign up for alerts
  • Open access
  • Published: 04 June 2024

HeteroTCR: A heterogeneous graph neural network-based method for predicting peptide-TCR interaction

  • Zilan Yu   ORCID: 1 , 2   na1 ,
  • Mengnan Jiang 1   na1 &
  • Xun Lan   ORCID: 1 , 2 , 3 , 4  

Communications Biology volume  7 , Article number:  684 ( 2024 ) Cite this article

Metrics details

  • Bioinformatics
  • Computational models

Identifying interactions between T-cell receptors (TCRs) and immunogenic peptides holds profound implications across diverse research domains and clinical scenarios. Unsupervised clustering models (UCMs) cannot predict peptide-TCR binding directly, while supervised predictive models (SPMs) often face challenges in identifying antigens previously unencountered by the immune system or possessing limited TCR binding repertoires. Therefore, we propose HeteroTCR , an SPM based on Heterogeneous Graph Neural Network (GNN), to accurately predict peptide-TCR binding probabilities. HeteroTCR captures within-type (TCR-TCR or peptide-peptide) similarity information and between-type (peptide-TCR) interaction insights for predictions on unseen peptides and TCRs, surpassing limitations of existing SPMs. Our evaluation shows HeteroTCR outperforms state-of-the-art models on independent datasets. Ablation studies and visual interpretation underscore the Heterogeneous GNN module’s critical role in enhancing HeteroTCR’s performance by capturing pivotal binding process features. We further demonstrate the robustness and reliability of HeteroTCR through validation using single-cell datasets, aligning with the expectation that pMHC-TCR complexes with higher predicted binding probabilities correspond to increased binding fractions.


T cells are a vital component of the adaptive immune system. T-cell receptors (TCRs) present on the surface of T cells have the ability to recognize peptides presented by Major Histocompatibility Complex (MHC) derived from various sources such as host proteins, pathogens, or tumors 1 . A majority of TCRs consist of a pair of α - and β -chains. During the maturation of T cells, TCR undergoes random rearrangement of variable ( V ), diversity ( D ), and joining ( J ) gene segments 2 . The interaction between TCR and peptide-MHC (pMHC) complex is mainly defined by three complementarity-determining regions (CDRs) for each chain 3 . Since the β -chain contains V -, D -, and J genes with higher diversity and the CDR3 regions are the determinant of the specific recognition of peptide 4 , most studies have focused on CDR3 regions of TCR β -chain 5 .

Reliably predicting peptide-TCR binding solely from sequences is of great significance for cancer immunotherapy and vaccine development 6 . Such predictions also provide valuable insights into patients’ immunological history, including responses to immunization, infections, and vaccines, thus paving the way for personalized healthcare 7 . Biological experiments to screen numerous unfiltered candidate immunogenic peptides are expensive and time-consuming 8 . Therefore, developing an effective computational method to predict peptide-TCR interactions is pressing due to the intricate nature of the molecular-level process. Notably, a peptide can bind to different TCRs, and conversely, a TCR is able to recognize multiple peptides 9 .

Existing models for predicting peptide-TCR binding can be divided into two categories 10 : (1) Unsupervised clustering models (UCMs) group similar TCR sequences together into clusters based on their shared features, including GLIPH 11 , TCRdist 12 , GIANA 13 , and iSMART 14 ; (2) Supervised predictive models (SPMs) can be further classified into categorical epitope models and generic models. Categorical epitope models are designed to learn TCR patterns that are specific to a particular peptide by using antigen-specific labels 15 . Examples of such models are TCR-classifier 16 , TCRex 17 , TCRGP 18 , and DeepTCR 19 . In contrast, generic models, such as NetTCR-1.0 20 , NetTCR-2.0 5 , ERGO Long Short Term Memory (LSTM) and Auto-encoder (AE) 21 , DLpTCR 22 , ImRex 23 , and TITAN 7 , can predict interactions between any TCR and peptide, without being restricted to a specific peptide.

Though UCMs have shown promise in various applications, they are not directly applicable to recognizing peptide-TCR binding 24 . Additionally, categorical epitope models require training for each peptide or set of peptides in multiclass classification, demanding ample TCR training data binding to the same peptide 23 . Furthermore, these models are inherently incapable of predicting the interactions of peptides not included in the training set (unseen epitopes) 7 . Existing generic models predict interactions between any TCR and peptide, without being restricted to a specific peptide, but rely on concatenating two interacting sequences encoded with a BLOSUM matrix or physicochemical properties 23 . These approaches might not directly focus on the interaction problem, instead learning embeddings for individual interactors, resulting in diminished model performance 23 . Moreover, although these studies perform well on the testing set that includes the training set peptides or TCRs, they cannot reliably predict interactions for those not present in the training set 6 . Additionally, they lack generalizability to new data from diverse databases 22 , due to the extensive variation in TCR and peptide sequences 23 .

Graph neural networks (GNNs) have made remarkable strides in recent years, establishing themselves as essential tools for a range of graph-based applications. Notably, they have demonstrated success in predicting chemical stability 25 , forecasting protein solubility 26 , modeling protein–protein interaction prediction 27 , and exploring drug–target interactions 28 . Therefore, motivated by these successes in the use of GNNs, we introduce HeteroTCR (Fig.  1 ), a Heterogeneous Graph Neural Network based SPM that utilizes only CDR3β sequence information from TCRs and peptides to improve predictive accuracy for peptide-TCR recognition across various datasets. Firstly, we extract the max-pooling layers of the pre-trained Convolutional Neural Network (CNN) module as numeric embeddings of TCRs and peptides to input the model. Then, we employ Heterogeneous GNN module as the backbone of our model, allowing HeteroTCR to extract information on between-type (peptide-TCR) interaction and within-type (TCR-TCR or peptide-peptide) similarity from input embeddings, thereby learning the probability of binding for classification. Existing SPMs mainly rely on the interaction information extracted from first-order neighborhoods while omitting insights originating from higher-order neighborhoods. An inherent advantage for HeteroTCR is that it can integrate information from neighborhoods of different orders, including both interaction information and similarity information. Our evaluation demonstrates that HeteroTCR outperforms state-of-the-art models on multiple independent datasets in terms of area under the curve (AUC) of the receiver operating characteristic (ROC). Ablation studies and visualization substantiate the essential role played by the Heterogeneous GNN module in enhancing the performance of HeteroTCR. This module adeptly captures critical features that underlie the binding process, showcasing its capacity for generalization across datasets with varying spatial distributions. Furthermore, we demonstrate HeteroTCR’s robustness and reliability in single-cell datasets by correlating predicted probabilities with experimentally derived pMHC-T cell binding fractions.

figure 1

To begin, the amino acid sequences of TCRs and peptides are encoded into numeric embeddings by a pre-trained CNN module. Subsequently, the Heterogeneous GNN module is employed to capture information pertaining to interactions between different types (peptide-TCR) and similarities within the same type (TCR-TCR or peptide-peptide). A circle node represents a TCR, while a triangle node represents a peptide. The edge of the solid line represents an interaction between a TCR and a peptide, while the dotted line represents no interaction, and no connection represents unknown. It is important to note that when initializing the graph, the nature of the relationship between peptides and TCRs is unknown, and we can only rely on the input training pairs to identify the existence of a relationship. We get the weight matrix by iterating 1000 epochs of training to learn and adjust the nature of the relationship between the given TCR and peptide, whether it is a solid line or a dotted line. We use a set of aggregator functions to learn aggregate feature information from K th -order neighborhoods of a node, with a default K value of 3. As shown in the graphs with green and blue shading, k  = 1, k  = 2, and k  = 3 represent the node information aggregated from the first, second, and third-order neighborhoods of the circle and triangle nodes. In other word, a given peptide/TCR can learn information from its 1 st -order binding TCRs/peptides, while the 1 st -order TCR/peptide can learn information from its 1 st -order binding peptides/TCRs which are 2 nd -order neighborhoods of given peptide/TCR. Therefore, a given peptide/TCR can extract between-type (peptide-TCR) interaction information from the 1 st - and 3 rd -order TCRs/peptides, and obtain within-type (TCR-TCR or peptide-peptide) similarity information from the 2 nd -order peptides/TCRs. This is because TCRs/peptides with shared features tend to interact with the same peptide/TCR. This process continues until order K is reached. Finally, an MLP is constructed utilizing these two numeric vectors to ascertain the presence of an interaction between a TCR and a peptide. Overall, HeteroTCR is composed of three main components: a pre-trained CNN, a Heterogeneous GNN, and an MLP classifier. The numeric embeddings extracted from the pre-trained CNN module are utilized as inputs for the Heterogeneous GNN module, which enables the extraction of information from the sequences. Specifically, the Heterogeneous GNN initially creates a global graph based on the entire dataset, and then trains on each pair of peptide-TCR inputs.

Model architecture

Conceptually, HeteroTCR divides the prediction of peptide-TCR interactions into three steps (Fig.  1 ). Firstly, a pre-training strategy is adopted, wherein the encoding vectors of the max-pooling layers before classification in CNN module are extracted as numeric embeddings of TCRs and peptides. Secondly, Heterogeneous GNN module is utilized to extract information on between-type (peptide-TCR) interaction and within-type (TCR-TCR or peptide-peptide) similarity. This results in TCR vectors containing information about interacting peptides and similar TCRs, and peptide vectors containing information about interacting TCRs and similar peptides. Finally, a Multi-Layer Perceptron (MLP) is created on top of these two numeric vectors to identify whether an interaction exists between a TCR and a peptide.

To numerically embed TCR and peptide sequences, we adopt a pre-training strategy inspired by Montemurro et al 5 who propose that the ability of classification in CNN-based prediction models is driven by the representation in the max-pooled CNN layer. Therefore, we extract the numeric embeddings of TCRs and peptides from the max-pooled CNN layer of the pre-trained CNN module for subsequent module. First, we encode amino acid (AA) sequences of TCRs and peptides using the BLOSUM50 matrix 29 , representing each AA as the score for substituting it with all 20 AAs. Then, TCR sequences and peptide sequences are individually processed using different convolution filters with kernel sizes {1, 3, 5, 7, 9}. Each kernel size’s convolutional output is max-pooled, yielding a single vector with 160 entries (80 for each input sequence) concatenated from multiple feature vectors. This 160-dimensional vector is fed into an MLP, producing the probability of peptide-TCR binding (Methods). Finally, the max-pooling layers of the pre-trained CNN module are extracted as numeric embeddings of TCRs and peptides, ready for the subsequent module.

For interaction information and similarity information on TCRs and peptides, existing algorithms for predicting peptide-TCR interactions typically extract sequence information from peptides and TCRs separately and simply concatenate them for classification prediction. However, this approach may not directly focus on the interaction problem, but learn an embedding for the individual interactors, resulting in reduced model performance 23 . Moreover, a single TCR can bind to multiple peptides, and a single peptide can bind to numerous TCRs. TCRs (peptides) associated with the same peptide (TCR) may have similarities to some extent. These established conditions consequently enable the capture of within-type (TCR-TCR or peptide-peptide) similarity using sequence information. Therefore, we adopt Heterogeneous GNN, a feature extraction module that stores information about different types of entities in nodes and their different types of relations in edges, to capture the within-type similarity information. We consider the complete dataset as a global graph, featuring two node types: TCR and peptide, and two edge types: TCR binding to peptide and peptide binding to TCR. It is worth noting that the two edge types are set to be equal, ensuring a balanced interaction between TCRs and peptides.

The features of nodes in Heterogeneous GNN are numeric embeddings of TCRs and peptides obtained from the max-pooling layers of the pre-trained CNN module. A solid line represents an interaction between a TCR and a peptide, while the dotted line represents no interaction, and no connection represents an unknown status (Fig.  1 ). Messages are transmitted on the edges through the network, and the weight matrix is obtained through backpropagation from the final MLP classifier (Methods). We use a set of aggregator functions to learn aggregate feature information from K th -order neighborhoods of a node, with a default K value of 3. In other words, a given peptide/TCR can learn information from its 1 st -order binding TCRs/peptides, while the 1 st -order TCR/peptide can learn information from its 1st-order binding peptides/TCRs which are 2 nd -order neighborhoods of given peptide/TCR. Therefore, a given peptide/TCR can extract between-type (peptide-TCR) interaction information from the 1st- and 3rd-order TCRs/peptides, and obtain within-type (TCR-TCR or peptide-peptide) similarity information from the 2 nd -order peptides/TCRs. This is because TCRs/peptides with shared features tends to interact with the same peptide/TCR. This process continues until order K is reached. Therefore, Heterogeneous GNN can extract information on between-type (peptide-TCR) interaction and within-type (TCR-TCR or peptide-peptide) similarity, allowing the vectors to learn about the probability of a binding event.

Finally, we utilize the numeric vectors of TCRs and peptides to learn their pairing. We construct an MLP (Methods) based on the Heterogeneous GNN module’s outputs. The final MLP layer is a single neuron for predicting the interaction between a TCR and a peptide. HeteroTCR outputs a continuous variable between 0 and 1, reflecting the predicted binding strength. A value greater than or equal to 0.5 indicates that there is an interaction between a peptide and a TCR, while a value below 0.5 predicts no interaction.

Comparisons with published methods on the independent testing dataset

We assessed the generalizability of HeteroTCR and other published methods on pair-based data sets (Methods). Specifically, the five models under comparison include NetTCR-1.0, NetTCR-2.0, ERGO LSTM, ERGO AE, and DLpTCR. NetTCR-1.0 20 is based on a simple 1D convolutional neural network (CNN), integrating peptide and CDR3 β sequence information encoded with BLOSUM50 matrix for the prediction of peptide-TCR specificity. NetTCR-2.0 5 modifies the network structure on the basis of NetTCR-1.0 and uses paired CDR3 α / β data as input instead of CDR3 β information only. Here, we exclusively employed the CDR3 β single-input model from NetTCR-2.0 for a fair comparison, omitting the utilization of the paired CDR3 α / β dual-input model. ERGO 21 applies two encoding methods, Long Short Term Memory (LSTM) acceptor encoding, and Autoencoder-based encoding, for predicting TCR-peptide binding. DLpTCR 22 uses a variety of mixed encoding methods, including one-hot, PCP, and PCA encoding, and integrates FCN, LeNet-5 and ResNet-20 to predict the interaction between peptide and TCR.

The codes for the aforementioned models is sourced from their respective GitHub repositories. The models were retrained and evaluated on our dataset. We trained the models on the IEDB dataset 30 , which comprises of 76,348 peptide-TCR pairs, using 5-fold cross-validation. The dataset was randomly divided into five equal parts, and the models were evaluated in each part in a rotating manner (Methods). For independent testing, we used data collected from the VDJdb 31 , which contains peptide-TCR pairs with confidence scores ranging from 0 to 3, where higher scores indicate more reliable records. Peptide-TCR pairs present in the training data were excluded from the testing data. Throughout our analysis, the area under the curve (AUC) of the receiver operating characteristic (ROC) was used as the metric to estimate model performance. All experiments were performed with default parameter settings.

As shown in Fig.  2 , the AUC scores of all models generally increased with improved dataset quality. However, due to a limited number of data pairs in VDJdb with a confidence score greater than or equal to 3 (only 32 pairs), yielding statistically significant results became impractical, leading us to exclude them. Additionally, the AUCs of HeteroTCR reached 0.6658, 0.6712, and 0.7039 in VDJdb with a confidence score greater than or equal to 0, 1, and 2 respectively, surpassing those of other published models (paired t-test P-value between HeteroTCR’s AUC and that of any other model was <0.00001) (details in Supplementary Tables  1 , 2 ).

figure 2

All boxplots: The center line of each boxplot marks the sample median, and the box extends from the lower to the upper quartile. Each colored point represents one data point. Each boxplot represents the results of n = 5 independent experiments. Our results show that HeteroTCR outperforms tested methods, as indicated by a paired t-test P-value of <0.00001 for comparisons with any other model.

Four types of data splitting methods to evaluate the generalization ability of the model

To further demonstrate the generalizability of HeteroTCR, we evaluated models based on four different types of data splitting methods (Methods), namely pair-based, TCR-based, antigen-based, and strict-based data sets. For the pair-based data sets, the peptide-TCR pairs present in the training dataset were removed from the testing dataset. For TCR-based data sets, peptide-TCR pairs involving TCRs present in the training dataset are excluded from the testing dataset. Similarly, for antigen-based data sets, peptide-TCR pairs involving peptides appearing in the training dataset are excluded from the testing dataset. In the case of strict-based data sets, peptide-TCR pairs involving either TCRs or peptides present in the training dataset are excluded from the testing dataset.

In this analysis, we trained models on IEDB data using four types of data splitting methods, and took McPAS-TCR 32 as the validation dataset for parameter selection. The evaluation was conducted on VDJdb with varying confidence scores (Methods). Importantly, the validation dataset was completely independent from the training set, and likewise, the testing dataset was also independent from both training and validation sets. This approach confirmed the model’s grasp of essential peptide-TCR binding properties, rather than confounding factors within the database.

As shown in Fig.  3 (performance comparison with other VDJdb confidence scores are detailed in Supplementary Figs.  1 – 3 ), we observed a general decrease in AUC scores across all models as data splitting methods became stricter. This decline in performance could stem from potential violations of the assumption of independent and identically distributed (i.i.d.) data across training, validation, and testing sets 7 . Additionally, stricter data splitting methods might lead to weaker data leakage and consequent performance degradation. Given the sparsity of current datasets, it is challenging for models to generalize to unseen epitopes. Notably, all models exhibited substantial performance drops on antigen-based and strict-based data sets as anticipated.

figure 3

For pair-based data sets, the validation dataset McPAS-TCR consists of 1574 pairs, and the testing dataset consists of 9206 pairs. For TCR-based data sets, antigen-based data sets, and strict-based data sets, the number of pairs in McPAS-TCR is 1436, 836, and 784, respectively, while VDJdb contains 7374, 1600, and 1208 pairs, respectively. All boxplots: The center line of each boxplot marks the sample median, the colored points scattered along each boxplot represent all the actual data points, and the box extends from the lower to upper quartile. Each boxplot represents the results of n  = 5 independent experiments. The paired t-test P-value between the AUC of HeteroTCR and that of any other model is < 0.00001, indicating that HeteroTCR performs significantly better than other methods.

HeteroTCR significantly outperformed other models across all dataset splitting with varying VDJdb confidence scores (paired t-test P-value between the AUC of HeteroTCR and that of any other model was <0.00001) (see Supplementary Tables  3 , 4 for more details). Although generalizability on strict-based data sets remained limited, HeteroTCR’s performance (AUC = 0.6535) surpassed random guessing (AUC = 0.5). These findings indicated HeteroTCR’s ability to learn from existing datasets and extrapolate predictions to unseen epitopes and TCRs.

Additionally, the lower variance in AUC for HeteroTCR compared to other methods was likely attributed to the robustness of the GNN model, which consistently performs well across diverse datasets and demonstrates a more stable predictive performance. This stability is a noteworthy characteristic that enhances the reliability and consistency of HeteroTCR’s predictive capabilities.

Comparisons to the state-of-the-art models

ImRex 23 and TITAN 7 are two state-of-the-art (SOTA) models for predicting unseen epitopes and TCRs on independent datasets. ImRex relies on an interaction map, which combines the physicochemical properties of both interactors on the amino-acid level, to predict peptide-TCR recognition under different dataset settings. TITAN exploits convolutions with an interpretable attention mechanism to aggregate local information and integrates the modalities, from which binding probabilities are predicted. To compare HeteroTCR directly, we trained and tested it under their respective training and testing data.

In ImRex (Table  1 ), the VDJdb dataset was used for training with 5-fold cross-validation, and the McPAS-TCR dataset served as an independent testing set, which was split into two subsets: one containing peptides already present in the training set (shared-epitope data), and the other containing peptides not seen during ImRex training (unseen-epitope data, which was referred as the unique-epitope data by the authors of ImRex 23 ). Negative interaction pairs were generated using a shuffling approach for each subset. We observed that HeteroTCR consistently outperformed ImRex on both independent testing sets.

For comparison with TITAN (Table  2 ), two different data splitting methods were utilized to evaluate model generalizability using 10-fold cross-validation. The TCR split ensured each TCR appeared in only one fold, avoiding overlap between validation and training datasets. The strict split extended this approach to both TCRs and peptides, ensuring that validation data were unseen during training. To ensure the separation of TCRs and peptides in their folds, negative data were generated by shuffling within each fold.

We found that HeteroTCR, utilizing only CDR3 sequences of TCRs and AA sequences of peptides, outperformed TITAN’s best model, which was pretrained on BindingDB 33 and fine-tuned on training data with full TCR sequences and SMILES encodings 34 of peptides, featuring a frozen epitope input channel with augmentations to enrich the data. HeteroTCR exhibited improved performance on unseen epitopes and TCRs with the strict data-splitting method.

The Heterogeneous GNN module is essential for the improved performance of HeteroTCR

To emphasize the advantages of HeteroTCR, we conducted ablation studies by establishing a baseline model. In computational terms, ablation studies refer to the systematic removal of components from the computational model to assess its impact on overall performance, rather than modifying predicted residues of peptides or TCRs. In our baseline model, we excluded the Heterogeneous GNN module, enabling us to assess the module’s contribution to enhancing both between-type (peptide-TCR) interaction and within-type (TCR-TCR or peptide-peptide) similarity. Models were trained on IEDB data using four types of data splitting methods, with McPAS-TCR as the validation dataset. Further evaluation took place independently on VDJdb with a confidence score greater than or equal to 0.

The AUC of HeteroTCR markly surpassed that of the baseline model (Fig.  4 ), indicating that the information on between-type interaction and within-type similarity learned by Heterogeneous GNN greatly enhanced the model’s performance (further details can be found in Supplementary Tables  5 and 6 ).

figure 4

All boxplots: The center line of each boxplot marks the sample median, the colored points scattered along each boxplot represent all the actual data points, and the box extends from the lower to upper quartile. Each boxplot represents the results of n  = 5 independent experiments. T-test P-value between HeteroTCR and the baseline model is <0.001.

Furthermore, we investigated the relationship between model performance and the parameter K (Supplementary Table  7 ). We considered K values ranging from 1 to 6. Our findings indicate a non-linear relationship between K and model performance, characterized by an initial improvement followed by a subsequent decline. Notably, the model achieves the highest AUC at K  = 4. However, considering the associated increase in training depth and computational cost at higher K values, we recommend adopting K  = 3 as it provides a balanced trade-off between time complexity and performance.

An inherent challenge in generic peptide-TCR models is their potential to memorize TCR motifs regardless of the peptide partner, leading to an inability to capture the true underlying molecular forces governing the binding process and merely managing to memorize spurious motifs present within the TCR training data 23 . To assess whether HeteroTCR captures essential binding features or memorizes spurious TCR motifs, we visualized model-extracted representations of peptide-TCR pairs in a 2-dimensional space using t-distributed stochastic neighbor embedding (t-SNE) 35 (Fig.  5a, b ). Due to the large number of peptide-TCR pairs in VDJdb with a confidence score greater than or equal to 0, we selected representative visuals from VDJdb with a confidence score greater than or equal to 1. Notably, using numeric representations extracted by the full model, cognate TCRs for a specific antigen clustered more tightly compared to a model without the Heterogeneous GNN module. Additionally, with the Heterogeneous GNN module, we observed a better separation between the orange dots (representing positive samples, indicating peptide-TCR interactions) and the purple dots (indicating negative samples, denoting no peptide-TCR interactions).

figure 5

a , b The t-SNE plot for the peptide-TCR pairs of the given peptide with and without a Heterogeneous GNN module. Each point represents a pair of peptide-TCR and is colored purple or orange, where purple represents no interaction and orange represents an interaction. Without Heterogeneous GNN module, the colored dots are scattered, and no clear boundary can be observed between purple and orange dots. In contrast, with Heterogeneous GNN, the colored dots are more clearly clustered and classified. c The degree of aggregation of the colored points (peptides with ≥15 cognate TCRs) is assessed by computing the center of the point set and the average distance of each point to the center. All boxplots: The center line of each boxplot marks the sample median, the colored points scattered along each boxplot represent all the actual data points, and the box extends from the lower to upper quartile. Each boxplot represents the results of n  = 5 independent experiments. d The per-peptide (with ≥15 cognate TCRs) AUC for the models with and without Heterogeneous GNN module, which are trained five times on IEDB and tested on VDJdb with confidence score greater than or equal to 1, based on the pair-based setting. All boxplots: The center line of each boxplot marks the sample median, the colored points scattered along each boxplot represent all the actual data points, and the box extends from the lower to upper quartile. Each boxplot represents the results of n  = 5 independent experiments.

To measure the degree of aggregation of positive and negative samples for the same antigen, we calculated the center of the point set and the average distance of each point to the center (Fig.  5c ). We found that the degree of aggregation of the samples with the Heterogeneous GNN module, is smaller than that without the Heterogeneous GNN module, indicating that cognate TCRs for a specific antigen cluster more tightly (paired t-test P-value between the models with and without Heterogeneous GNN module is \(3.9\times {10}^{-10}\) ) (additional details in Supplementary Table  8 ). Figure  5d shows the AUC in per-peptide classification of positive and negative samples (≥15 cognate TCRs), demonstrating the Heterogeneous GNN module’s reasonable performance (paired t-test P-value between the models with and without Heterogeneous GNN module is \(1.2\times {10}^{-7}\) ) (additional details in Supplementary Table  9 ). This observation reveals that the information on between-type (peptide-TCR) interactions and within-type (TCR-TCR or peptide-peptide) similarity learned by HeteroTCR prevents reliance on TCR motif patterns for interaction prediction.

TCRs recognizing the same antigen can exhibit diverse amino acid sequences. To investigate the impact of TCR sequence diversity on HeteroTCR, we evaluated model performance using TCRs with decreasing similarity within the training data (Fig.  6a ), and trained with the full data but tested using TCRs with decreasing similarity to the training data (Fig.  6b ). TCR similarity was measured using the mean Levenshtein ratio for each TCR to all TCRs in the training data (Methods). A higher mean Levenshtein ratio suggests a TCR sequence is closer to the TCRs in the training data. AUC analysis was conducted for TCRs below specified Levenshtein ratio cut-offs (Methods). AUC scores are shown for model trained (Fig.  6a ) or tested (Fig.  6b ) with TCRs above each cut-off. The outcomes reveal HeteroTCR’s relative robust as TCR similarity levels decrease. With diminishing TCR similarity within the training set, HeteroTCR’s performance slightly drops (Fig.  6a ), indicating a subtle influence of TCR motifs on the model. As the similarity between testing and the training TCRs (Fig.  6b ) wanes, AUC scores remain steady, indicating HeteroTCR’s capability to generalize across diverse TCR motifs.

figure 6

a Performance of HeteroTCR with decreasing similarity of TCRs within the training data. The abscissa represents the TCR Levenshtein ratio, with higher values indicating greater similarity. The blue bars represent AUC scores (left ordinate), and the red line indicates the number of training data points (right ordinate) for each cut-off. The cut-offs used are 0.30, 0.29, 0.28, 0.27, 0.26, 0.25, and 0.24. b Performance of HeteroTCR with decreasing similarity of TCRs between the training and testing data. The abscissa represents the TCR Levenshtein ratio, where higher values indicate that testing TCRs are more similar to the training TCRs. The blue bars represent AUC scores (left ordinate), and the red line indicates the number of testing data points (right ordinate) for each cut-off. The cut-offs used are 0.30, 0.29, 0.28, 0.27, 0.26, 0.25, and 0.24, corresponding to thresholds where similar TCR data points are progressively removed from the testing dataset.

Additionally, we investigated the impact of the number of peptides on model performance (Fig.  7 ). In this analysis, we trained models on the IEDB dataset using strict-based datasets and utilized McPAS-TCR as the validation dataset for parameter selection. The evaluation was performed on VDJdb with a confidence score greater than or equal to 0. In the experiments, the initial training dataset consisted of 559 unique peptides. We randomly subsampled them to reduce the quantity to 100, 200, 300, 400, and 500. Our observations revealed that the model performance continued to improve with increased input data, suggesting that HeteroTCR benefited from the inclusion of more peptides (additional details in Supplementary Table  10 ). In summary, our experiments demonstrate that increasing the number of peptides in the training dataset consistently improves the performance of the model, underscoring the importance of diverse antigenic exposures in enhancing predictive accuracy.

figure 7

The shaded area depicts the mean ± the standard deviation over five repeated experiments. The blue points represent the mean data points, while the grey points ( n  = 5) scattered around each mean point represent all the actual data points. In the validation dataset and testing dataset, the number of pairs for different numbers of subsampled peptides is 784 and 1208, respectively. For different numbers of subsampled peptides (ranging from 100 to 559), the corresponding numbers of pairs in the training dataset are 7180, 22440, 40932, 52636, 69816, 76348, respectively.

HeteroTCR binding predictions correlate with experimentally derived pMHC-T cell binding fractions

To assess the performance of the model from an alternative angle, we investigated whether the predictions of the HeteroTCR model correlate with experimentally derived pMHC-T cell binding fractions (Methods). The data utilized in our study was generated from the 10x Genomics Chromium Single Cell Immune Profiling platform 36 (Methods). We analyzed single-cell datasets containing profiles of CD8 + T cells specific to 44 different pMHC multimers, sourced from four healthy donors. The binding specificity between each T-cell and tested pMHC was quantified by counting the number of unique molecular identifier (UMI) sequences associated with each specific pMHC in the T-cell. After data curation and the computation of binding fractions (Methods), we proceeded to calculate the Spearman correlation coefficient between HeteroTCR-predicted binding probabilities and binding fractions, i.e. the fraction of a T cell clone bound to a specific TCR.

Figure  8a–d illustrates the correlation between the predicted binding probabilities of HeteroTCR and binding fractions. Notably, the dataset includes cases where the binding fraction is 0, illustrating an absence of binding specificity. A positive correlation observed in the figures indicates the association between the model’s predicted probabilities and the binding events. In Fig.  8e–h , the correlation between the two variables in data excluding instances where the binding fraction equals 0. Consequently, the figures exclusively comprise data with binding specificity, and a positive correlation depicts that the predicted binding probabilities of HeteroTCR are associated with binding strength in instances where binding occurs.

figure 8

a – d Binding predictions for donors 1–4 correlate with binding fractions, where binding fractions include cases equal to 0. The number of data for each donor is 29495, 62956, 41496, and 13430, respectively. e – h Binding predictions for donors 1–4 correlate with binding fractions, where binding fractions exclude cases equal to 0. The number of data for each donor is 9909, 24156, 13114, and 4792, respectively.

In this study, we introduce HeteroTCR, a Heterogeneous Graph Neural Network method leveraging the sequence information form the peptide and CDR3 β region of TCR for predicting peptide-TCR binding probability. HeteroTCR is designed to extract both information on between-type (peptide-TCR) interaction and within-type (TCR-TCR or peptide-peptide) similarity from input sequence. As far as we know, HeteroTCR is the first model to apply graph neural networks to peptide-TCR interaction prediction tasks. Our evaluation demonstrated that HeteroTCR outperforms existing methods in predicting peptide-TCR binding probabilities, particularly across various independent testing datasets and dataset settings. Although the generalization capability of HeteroTCR on strict-based data sets is limited, its overall performance remains robust. These results highlight HeteroTCR’s enhanced reliability and robustness in predicting immunogenic peptides recognized by TCR. The HeteroTCR program, trained model parameters, machine learning platform and hardware (Methods), and datasets are available on GitHub ( ).

To further analyze the superiority of HeteroTCR design, we establish a baseline model for ablation studies. Our results show that the information on between-type interaction and within-type similarity learned by the Heterogeneous GNN module enhances model performance considerably. Utilizing t-SNE visualization on numeric vectors processed with and without the Heterogeneous GNN module, we uncover HeteroTCR’s ability to capture fundamental molecular binding forces rather than memorizing TCR pseudo motifs. Additionally, we investigate the impact of TCR pattern similarity on HeteroTCR by varying Levenshtein ratio cut-offs on training and testing sets, illustrating that HeteroTCR’s strong performance is partly contributed by within-type similarity.

Moreover, we investigated the correlation between HeteroTCR’s binding predictions and experimentally derived pMHC-T cell binding fractions using publicly available 10x Genomics data from single cell droplet sequences in the presence of DNA barcoded pMHC dextramers. While the model is promising in aiding the identification of immunogenic epitopes, a wide range of binding strengths were observed even among TCR with the highest HeteroTCR GNN-inferred interaction scores. This underscores that HeteroTCR may have the utility to prioritize candidate TCRs of clinical relevance, but binding prediction will require experimental validation.

Heterogeneous GNN, integrating diverse node and edge information 37 , opens promising avenues for future research to expand on HeteroTCR. Additional information, including the alpha chains of TCRs, V / J genes, other CDR loops, and the MHC contexts, can be incorporated into the graph as distinct node types with different vector dimensions and embedding methods. Edges can also have weights attached to represent different degrees of importance and the proportion of message transmission. Notably, HeteroTCR has important implications for designing TCR-T or neoantigen vaccine therapies, and offers a modeling framework for various molecular interactions like protein-protein or antibody-antigen interactions.

Here, we collected data from three well-known databases: IEDB 30 , VDJdb 31 , and McPAS-TCR 32 . The Immune Epitope Database (IEDB) is a freely available resource that contains a comprehensive collection of antibody and T cell epitopes in humans, non-human primates, and other animal species in the context of infectious disease, allergy, autoimmunity, and transplantation. The IEDB dataset was downloaded from on November 17, 2021. The VDJdb is a curated database of TCR sequences with known antigen specificities and confidence scores to highlight the most reliable records. The McPAS-TCR is a manually curated catalog that links TCR sequences to their antigen targets or to the pathology and organ with which they are associated. The VDJdb and McPAS-TCR were collected from and , both on June 12, 2021. To construct reliable datasets, we took five preprocessing steps on the raw data.

Step 1. filtering data: (i) The data used in this study were restricted to TCR β chain sequences due to the scarcity of paired chain data. (ii) Peptide-CDR3 β pairs were adopted because the complementarity-determining region 3 (CDR3) is the key determinant of specificity in antigen recognition. (iii) Given that peptide with lengths ranging from 8 to 15 amino acids (AAs) and CDR3 β sequences containing 10–20 AAs account for the majority of datasets 22 , we focused on the CDR3 β sequences with 10–20 AAs and peptides with 8–15 AAs presented by human MHC class I molecules, as they are likely to represent the most common and biologically relevant cases.

Step 2. removing invalid sequences: Since the majority of CDR3 β chains begin with the amino acid ‘C’ and end with ‘F’ 38 , we excluded peptide-CDR3 β pairs whose CDR3 β sequences do not adhere to this pattern. Furthermore, we eliminated peptide-CDR3 β sequences containing lowercase letters or letters that are not amino acids as these likely result from errors during database collection and collation.

Step 3. removing redundant sequence pairs: Duplicate sequence pairs may exist in the databases due to the omission of information such as the CDR1 and CDR2 regions and HLA molecules. To address this, we removed redundant data within each database by retaining only unique peptide-CDR3 β sequence pairs.

Step 4. screening antigen-specific TCR CDR3s: iSMART 14 is a computational method that performs a specially parameterized pairwise local alignment on TCR CDR3 sequences to group them into antigen-specific clusters with high efficiency. Since TCRs usually share conserved sequence features when recognizing the same pMHC epitope 22 , we employed iSMART as a filter to identify and cluster antigen-specific TCR CDR3s, removing the CDR3s that are not selected in the cluster.

Step 5. constructing datasets: The peptide-CDR3 pairs processed in Steps 1–4 can be added to the set of interacting pairs of CDR3s and peptides. However, to train supervised models, both positive and negative examples, which are pairs of CDR3s and peptides that do not recognize each other, are required. To generate negative data, each CDR3 sequence in the positive dataset was paired with a peptide that has not been shown to interact with the corresponding CDR3 by shuffling or mismatching the sequences. The enormous diversity of potential TCR and peptide sequences makes it unlikely for a randomly selected TCR to bind to a specific peptide, thereby making the shuffling method feasible 39 . In this study, the shuffling method was used instead of sampling uniformly from healthy CDR3 reference repertoire sets, as using an external negative reference dataset may introduce an inherent bias that may result in over-optimistic performance 23 .

In the Results 2.2 section, we used IEDB dataset as the training dataset with 5-fold cross-validation to generate five optimal parameter models, and evaluated these five models on an independent testing dataset (VDJdb) to get five AUC metrics.

In the Results 2.3 section, we trained the model on the data from IEDB dataset, took McPAS-TCR as the validation dataset to select the optimal parameter model, and evaluated the model on VDJdb. We repeated this process five times, so we got five optimal parameter models each time and five AUC metrics on the independent testing dataset.

In the Results 2.4 section, to avoid being accused of manipulating data using complex filtering steps, we directly used the data of SOTA models to train and evaluate HeteroTCR and compared the results in their papers.

Four types of data splitting methods

In order to further refine our experiment and show the performance of HeteroTCR, we redefined the data splitting methods into four types:

Pair-based data sets/seen epitopes and seen TCRs.

TCR-based data sets/unseen TCRs.

Antigen-based data sets/unseen epitopes.

Strict-based data sets/unseen epitopes and unseen TCRs.

For the pair-based data sets, it simply guarantees that the interacting peptide-TCR pairs in the testing dataset are not contained in the training dataset. It is worth noting that the peptide or TCR alone in the training dataset may exist in the testing dataset. This division method has been widely adopted by most models.

For the TCR-based data sets, if a TCR appears in the training dataset, it cannot be present in the testing dataset.

For the antigen-based data sets, the peptide that appears in the training dataset will no longer appear in the testing dataset.

For the strict-based data sets, it is necessary to ensure that both the TCR and the peptide in the training dataset do not appear in the testing dataset. This approach differs from the pair-based data sets, where the model’s generalization ability is evaluated by predicting whether TCRs or peptides that are not in the database can stably bind.

In addition to the above four types of data splitting methods, our model utilized cross-validation across different datasets to demonstrate that our model learns essential information about TCRs and peptides, rather than simply memorizing the specifics of a single database. More details regarding the datasets can be found in the results section.

Embedding of TCRs and peptides

Firstly, the TCR sequences were zero-padded to a maximum length of 20, while the peptide sequences to a maximum length of 15. We then encoded AA sequences of TCRs and peptides using the BLOSUM50 matrix, in which each AA is represented as the score for substituting the AA with all the 20 amino AAs. Hence, the BLOSUM encoding scheme maps a sequence with a length of l to an array with a size of l × 20. Secondly, the TCR sequences and the peptide sequences were separately deconvoluted by different convolution kernels with kernel size {1, 3, 5, 7, 9}, in which different features were integrated through different convolution kernels to filter the whole sequence. For each kernel size, the convolutional output was max-pooled and the resulting feature vectors were concatenated in a single vector with 160 entries (80 for each input sequence). Thirdly, the 160-dimensional vector was fed into a dense layer of 32 hidden neurons and the output consists of one single neuron, giving the probability of a peptide-TCR pair to bind (Fig.  1 ). The activation function used through the network is the sigmoid function. The model was trained for 1000 epochs with early stopping and patience of 50 epochs. The weights were updated using the Adam optimizer with a learning rate of 0.001. The batch size is 512 and the loss function is binary cross-entropy. Finally, the 160-dimensional max-pooling layers of the pre-trained CNN module were extracted as numeric embeddings of TCRs and peptides to input the next step.

Heterogeneous GNN module

Firstly, we regard the entire dataset as a global graph. A circle node represents a TCR, while a triangle node represents a peptide (Fig.  1 ). The edge of the solid line (label 1) represents an interaction between a TCR and a peptide, while the dotted line (label 0) represents no interaction, and no connection represents unknown. The difference between solid and dotted lines is only used for the calculation of loss function and backpropagation of the MLP in the training dataset, as well as the evaluating the model in the testing dataset, and it does not participate in any other process to cause data leakage. In other words, when initializing the graph, the nature of the relationship between peptides and TCRs is unknown, and we can only rely on the input training pairs to identify the existence of a relationship. We get the weight matrix by iterating 1000 epochs of training to learn and adjust the nature of the relationship between the given TCR and peptide, whether it is a solid line or a dotted line.

Secondly, the idea behind Heterogeneous GNN can be broadly divided into two steps. We use a set of aggregator functions to learn aggregate feature information from K th -order neighborhood of a node, with a default K value of 3. The activation function used through Heterogeneous GNN is the Leaky ReLUs function. The input of Heterogeneous GNN is an 80-dimensional numeric embedding of TCR and an 80-dimensional numeric embedding of peptide, and the output is a 1024-dimensional TCR vector and a 1024-dimensional peptide vector.

Let \(T=\{{t}_{1},{t}_{2},\ldots ,{t}_{m}\}(\left|T\right|=m)\) and \(P=\{{p}_{1},{p}_{2},\ldots ,{p}_{n}\}(\left|P\right|=n)\) denote the set of TCRs and peptides, respectively. Let \({I}^{+}=\{{y}_{t,p}{|t}\in T,{p}\in P\}\) denote the positive interactions, where \({y}_{t,p}\) indicates that TCR t has interacted with peptide p . Let \({I}^{-}=\{{y}_{t,p}{|t}\in T,{p}\in P\}\) denote the negative interactions, where \({y}_{t,p}\) indicates that TCR t has no interacted with peptide p . Moreover, a peptide-TCR interaction graph is constructed, denoted as \({{{{{\mathscr{G}}}}}}({{{{{\mathscr{V}}}}}}{{{{{\mathscr{,}}}}}}{{{{{\mathscr{E}}}}}})\) , where \({{{{{\mathscr{V}}}}}}{{{{{\mathscr{=}}}}}}T\cup P\) is the set of nodes and \({{{{{\mathscr{E}}}}}}=\{(t,p)|{y}_{t,p}\in I,t\in T,p\in P,I={I}^{+}\cup {I}^{-}\}\) is the edge set.

More formally, for the Graph \({{{{{\mathscr{G}}}}}}({{{{{\mathscr{V}}}}}}{{{{{\mathscr{,}}}}}}{{{{{\mathscr{E}}}}}})\) , the input features of TCRs or peptides are \(\left\{{x}_{v},\,\forall v\in {{{{{\mathscr{V}}}}}}\right\}\) , and the edges \(\left\{\forall e{{{{{\mathscr{\in }}}}}}{{{{{\mathscr{E}}}}}}\right\}\) represent the relationships between TCRs and peptides, whether positive or negative. The TCR or peptide embedding \({z}_{v}\) generation algorithm is described as follows:

where \({h}_{v}^{k}\) is the representation of node v at k-th order. \({h}_{v}^{0}\) is initialized by the input features of TCRs or peptides. \({{{{{\mathscr{N}}}}}}(v)\) denotes the neighbors of node v . \({z}_{v}\) is final representations of node v . \({W}_{k}\) and \({B}_{k}\) are trainable weight matrices, the depth K is 3 by default and the non-linearity σ is Leaky ReLU. The differentiable aggregator functions AGGREGATE is the average of neighbor’s previous layer embeddings. \({h}_{{{{{{\mathscr{N}}}}}}(v)}^{k}\) is trivially computed by \({h}_{{{{{{\mathscr{N}}}}}}(v)}^{k}={\sum}_{u\in {{{{{\mathscr{N}}}}}}(v)}\frac{{h}_{u}^{k-1}}{\left|{{{{{\mathscr{N}}}}}}(v)\right|}\) with |·| donating the number of the node v neighbors. To generate the representation of node v , it first aggregates the representations of \({{{{{\mathscr{N}}}}}}(v)\) and then updates the representation of v by concatenating its representation at (k-1)-th order and the aggregated representations.

MLP classifier

A linear layer transforms one vector into another vector, which refers to a single-layer neural network without hidden layers. The MLP is a stack of linear layers with hidden layers. For classification tasks, HeteroTCR uses an MLP with two hidden layers (Fig.  1 ). The two numerical vectors (2048 dimension) from the Heterogeneous GNN module are concatenated into a single layer with 512 neurons activated by ReLU, followed by a dense layer with 256 neurons also activated by ReLU, and the last layer with a single neuron with sigmoid activation. Mathematically, the output is a continuous variable between 0 and 1, representing the predicted binding strength between TCR and peptide. HeteroTCR was trained for 1000 epochs with 512 batch sizes using the Adam optimizer with a learning rate of 0.0001 and a weight decay of 0 by default. The loss function is binary cross-entropy.

Levenshtein similarity ratio

Here, the Levenshtein ratio was used as a measure of the similarity between TCR sequences. The Levenshtein ratio is based on the ldist , which is not Levenshtein distance, but the sum of the costs. Given two strings, the ldist describes the count of modifications needed to transform one word into another. The possible changes are deletion, insertion, and replacement. The count of deletion and insertion is 1, while the count of replacement is 2. The Levenshtein ratio is given by the formula

where u and v represent two TCR sequences, and |∙| defines their length.

We defined the Levenshtein similarity ratio of a TCR as the average of the Levenshtein ratio from the TCR to each TCR in the training set, and then removed the data exceeding the specified ratio cut-offs. In Fig.  6a , we calculated the average of each TCR’s Levenshtein ratio in the training set to all of the training TCRs and thus reduced TCR similarity within the training set. In Fig.  6b we calculated the average of each TCR’s Levenshtein ratio in the testing set to all of the training TCRs and thus reduced the TCR similarity between the testing set and the training set.

Data curation of the 10× Genomics platform and calculation of binding fractions

The data utilized in our study was generated from the 10× Genomics Chromium Single Cell Immune Profiling platform ( ). The raw data is accessible for download at . Closely examining four single-cell datasets, we analyzed profiles of CD8 + T cells specific to a highly multiplexed panel consisting of 44 different pMHC multimers, alongside 6 negative control pMHC multimers, sourced from four healthy donors. The binding specificity between T cells and each pMHC complex was quantified using the UMI counts as a binding indicator.

In general, a T cell typically expresses only one pair of functional TCRs, and thus, we selectively retain clones of T cells expressing a single pair of TCR α and TCR β chains. We focused only on expanded clones, as the UMI counts of each cell were inherently noisy due to dropouts and high variances in single-cell experiments. Consequently, we opted to utilize the binding fraction of each T cell clone with the same TCR as a measure of antigen affinity to a TCR. The binding fraction of a clone is determined by the following formula:

where m represents the number of T cells within the clone exhibiting a higher UMI count for a given antigen compared to the maximum UMI count among the 6 negative control antigens, and n denotes the clone size, i.e., the number of T cells with identical TCRs. It is noteworthy that instances where m equals 0 are only considered in the following two scenarios: when the UMI count of T cells within the clone for a given antigen is less than the maximum UMI count among the 6 negative control antigens, or when the UMI count of T cells within the clone for a given antigen equals the maximum UMI count among the 6 negative control antigens, and both are non-zero. Specifically, we do not consider cases where m equals 0 when the UMI count of T cells within the clone for a given antigen is 0, and the maximum UMI count among the 6 negative control antigens is also 0. This omission is due to the sparse nature of the original data, characterized by numerous zero values, leading to excessive noise and inclusion of many meaningless data points. Finally, we calculated the Spearman correlation coefficient between predicted binding probabilities and binding fractions.

Machine learning platform and hardware

The pre-trained CNN module was implemented with Keras 2.6.0 ( ) using the Tensorflow backend and Python 3.7.0. The Heterogeneous GNN module and the MLP classifier were implemented with PyTorch 1.9.1 ( ) using Python 3.7.11. PyG (PyTorch Geometric) 37 , a library built upon PyTorch to easily write and train GNNs, was employed for modeling and processing graph-structured data. The metric evaluation was implemented with TorchMetrics package 40 . All deep learning models were trained on a single NVIDIA A40 graphics card.

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

All additional data that support the conclusions from this manuscript are available in the Supplementary Information. Source data and codes for the main figures, as well as all data utilized for HeteroTCR training and evaluation, are shared on the github repository ( ) and its Zenodo ( ). The dataset used for training and testing were collected from IEDB , VDJdb ( ) and McPAS-TCR ( ). The detailed information of the 10x Genomics cohort is available at . The raw data for single-cell datasets is available at .

Code availability

HeteroTCR is available on github ( ) and its Zenodo ( ), together with a usage documentation and several example testing datasets.

Szeto, C., Lobos, C. A., Nguyen, A. T. & Gras, S. TCR recognition of Peptide–MHC-I: Rule makers and breakers. Int. J. Mol. Sci. 22 , 68 (2020).

Article   PubMed   PubMed Central   Google Scholar  

Lefranc, M.-P. Nomenclature of the human T cell receptor genes. Curr. Protoc. Immunol . 40 , A.1O.1-A.1O.23 (2000).

Yu, K., Shi, J., Lu, D. & Yang, Q. Comparative analysis of CDR 3 regions in paired human αβ CD 8 T cells. FEBS Open Bio 9 , 1450–1459 (2019).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Rossjohn, J. et al. T cell antigen receptor recognition of antigen-presenting molecules. Annu. Rev. Immunol. 33 , 169–200 (2015).

Article   CAS   PubMed   Google Scholar  

Montemurro, A. et al. NetTCR-2.0 enables accurate prediction of TCR-peptide binding by using paired TCRα and β sequence data. Commun. Biol. 4 , 1060 (2021).

Joglekar, A. V. T cell antigen discovery. Nat. Methods 18 , 8 (2021).

Article   Google Scholar  

Weber, A., Born, J. & Rodriguez Martínez, M. TITAN: T-cell receptor specificity prediction with bimodal attention networks. Bioinformatics 37 , i237–i244 (2021).

Holler, P. D. et al. In vitro evolution of a T cell receptor with high affinity for peptide/MHC. Proc. Natl Acad. Sci. 97 , 5387–5392 (2000).

Petrova, G., Ferrante, A. & Gorski, J. Cross-reactivity of T cells and its role in the immune system. Crit. Rev. Immunol. 32 , 349–372 (2012).

Hudson, D., Fernandes, R. A., Basham, M., Ogg, G. & Koohy, H. Can we predict T cell specificity with digital biology and machine learning? Nat. Rev. Immunol . . (2023).

Glanville, J. et al. Identifying specificity groups in the T cell receptor repertoire. Nature 547 , 94–98 (2017).

Dash, P. et al. Quantifiable predictive features define epitope-specific T cell receptor repertoires. Nature 547 , 89–93 (2017).

Zhang, H., Zhan, X. & Li, B. GIANA allows computationally-efficient TCR clustering and multi-disease repertoire classification by isometric transformation. Nat. Commun. 12 , 4699 (2021).

Zhang, H. et al. Investigation of antigen-specific T-cell receptor clusters in human cancers. Clin. Cancer Res. 26 , 1359–1371 (2020).

Mösch, A., Raffegerst, S., Weis, M., Schendel, D. J. & Frishman, D. Machine learning for cancer immunotherapies based on epitope recognition by T cell receptors. Front. Genet. 10 , 1141 (2019).

De Neuter, N. et al. On the feasibility of mining CD8+ T cell receptor patterns underlying immunogenic peptide recognition. Immunogenetics 70 , 159–168 (2018).

Article   PubMed   Google Scholar  

Gielis, S. et al. Detection of enriched T cell epitope specificity in full T cell receptor sequence repertoires. Front. Immunol. 10 , 2820 (2019).

Jokinen, E., Huuhtanen, J., Mustjoki, S., Heinonen, M. & Lähdesmäki, H. Predicting recognition between T cell receptors and epitopes with TCRGP. PLOS Comput. Biol. 17 , e1008814 (2021).

Sidhom, J.-W., Larman, H. B., Pardoll, D. M. & Baras, A. S. DeepTCR is a deep learning framework for revealing sequence concepts within T-cell repertoires. Nat. Commun. 12 , 1605 (2021).

Jurtz, V. I. et al. NetTCR: Sequence-based prediction of TCR Binding to Peptide-MHC complexes using convolutional neural networks. (2018).

Louzoun, Y. Prediction of specific TCR-peptide binding from large dictionaries of TCR-peptide Pairs. Front. Immunol. 11 , 10 (2020).

Google Scholar  

Xu, Z. et al. DLpTCR: an ensemble deep learning framework for predicting immunogenic peptide recognized by T cell receptor. Brief. Bioinform. 22 , bbab335 (2021).

Moris, P. et al. Current challenges for unseen-epitope TCR interaction prediction and a new perspective derived from image classification. Brief. Bioinform. 22 , bbaa318 (2021).

Gao, Y. et al. Pan-Peptide meta learning for T-cell receptor–antigen binding recognition. Nat. Mach. Intell. 5 , 236–249 (2023).

Li, X. et al. DeepChemStable: Chemical stability prediction with an attention-based graph convolution network. J. Chem. Inf. Model. 59 , 1044–1049 (2019).

Chen, J., Zheng, S., Zhao, H. & Yang, Y. Structure-aware protein solubility prediction from sequence through graph convolutional network and predicted contact map. J. Cheminform. 13 , 7 (2021).

Zhou, H., Wang, W., Jin, J., Zheng, Z. & Zhou, B. Graph neural network for protein-protein interaction prediction: a comparative study. Mol. Basel Switz . 27 , 6135 (2022).

Zhang, Z. et al. Graph neural network approaches for drug-target interactions. Curr. Opin. Struct. Biol. 73 , 102327 (2022).

Henikoff, S. & Henikoff, J. G. Amino acid substitution matrices from protein blocks. Proc. Natl Acad. Sci. 89 , 10915–10919 (1992).

Vita, R. et al. The Immune Epitope Database (IEDB): 2018 update. Nucleic Acids Res. 47 , D339–D343 (2019).

Bagaev, D. V. et al. VDJdb in 2019: database extension, new analysis infrastructure and a T-cell receptor motif compendium. Nucleic Acids Res. 48 , D1057–D1062 (2020).

Tickotsky, N., Sagiv, T., Prilusky, J., Shifrut, E. & Friedman, N. McPAS-TCR: a manually curated catalogue of pathology-associated T cell receptor sequences. Bioinformatics 33 , 2924–2929 (2017).

Gilson, M. K. et al. BindingDB in 2015: A public database for medicinal chemistry, computational chemistry and systems pharmacology. Nucleic Acids Res. 44 , D1045–D1053 (2016).

Weininger, D., Weininger, A. & Weininger, J. L. SMILES. 2. Algorithm for generation of unique SMILES notation. J. Chem. Inf. Comput. Sci. 29 , 97–101 (1989).

Article   CAS   Google Scholar  

van der Maaten, L. & Hinton, G. E. Visualizing Data using t-SNE. J. Mach. Learn. Res. 9 , 2579–2605 (2008).

Schwab. CD8+ T cells of healthy donors sorted for Dextramer positive cells. Zenodo (2022).

Fey, M. & Lenssen, J. E. Fast Graph Representation Learning with PyTorch Geometric. in ICLR Workshop on Representation Learning on Graphs and Manifolds (2019).

Tong, Y. et al. SETE: Sequence-based Ensemble learning approach for TCR Epitope binding prediction. Comput. Biol. Chem. 87 , 107281 (2020).

Fischer, D. S., Wu, Y., Schubert, B. & Theis, F. J. Predicting antigen specificity of single T cells based on TCR CDR 3 regions. Mol. Syst. Biol . 16 , e9416 (2020).

Detlefsen, N. S. et al. TorchMetrics - measuring reproducibility in PyTorch. J. Open Source Softw. 7 , 4101 (2022).

Download references


This work was funded by grants from the Tsinghua University Independent Research Project (ID: 52302102423), the Tsinghua University Spring Breeze Fund, and the Beijing Institute of Technology’s Proof of Concept Project for Tumor Neoantigen Personalized TCR-T Therapy.

Author information

These authors contributed equally: Zilan Yu, Mengnan Jiang.

Authors and Affiliations

School of Medicine, Tsinghua University, 100084, Beijing, China

Zilan Yu, Mengnan Jiang & Xun Lan

Centre for Life Sciences, Tsinghua University, 100084, Beijing, China

Zilan Yu & Xun Lan

Tsinghua-Peking Center for Life Sciences, MOE Key Laboratory of Tsinghua University, Beijing, China

MOE Key Laboratory of Bioinformatics, Tsinghua University, 100084, Beijing, China

You can also search for this author in PubMed   Google Scholar


Zilan Yu, Mengnan Jiang, and Xun Lan designed the framework of this study. Zilan Yu devised the model architecture and implemented the code. Mengnan Jiang contributed biological insights, created figures, and acquired and preprocessed data for analysis. Zilan Yu wrote the paper with the help of other authors. Xun Lan supervised the study. All authors read and approved the final paper.

Corresponding author

Correspondence to Xun Lan .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Communications Biology thanks Vy Nguyen, Xi Jiang, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Tobias Goris. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Peer review file, supplementary information, reporting summary, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit .

Reprints and permissions

About this article

Cite this article.

Yu, Z., Jiang, M. & Lan, X. HeteroTCR: A heterogeneous graph neural network-based method for predicting peptide-TCR interaction. Commun Biol 7 , 684 (2024).

Download citation

Received : 06 October 2023

Accepted : 23 May 2024

Published : 04 June 2024


Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

  • Explore articles by subject
  • Guide to authors
  • Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

research methods for communication

  • Weight Management
  • Nutrition Facts
  • Nutrition Basics
  • Meal Delivery Services
  • Fitness Gear
  • Apparel & Accessories
  • Recipe Nutrition Calculator
  • Weight Loss Calorie Goal
  • BMI Calculator
  • Body Fat Percentage Calculator
  • Calories Burned by Activity
  • Daily Calories Burned
  • Pace Calculator
  • Editorial Process
  • Meet Our Review Board

What Happens to Your Body When You Take NAD

Getty Images / PhotoAlto/Laurence Mouton

Understanding NAD

  • Potential Effects

Current Research

Dosage and delivery methods, natural ways to boost nad.

It's a typical weekday morning. Before the hustle and bustle of the day begins, you sip your collagen coffee while scrolling your social feeds. One after the other, almost every swipe is a post about the next best anti-aging elixir, compound, exercise, or food. Then, NAD pops up—a supplement touted by the likes of Hailey Bieber and Kendall Jenner as an anti-aging goldmine worthy of using for a lifetime. But is it really, though?

NAD—short for nicotinamide adenine dinucleotide—is a cellular coenzyme responsible for turning the food you eat into usable energy. It also plays a major role in DNA repair, cellular processes, and immune system function. Even though the body produces NAD naturally, its production of NAD declines as you age.

That means the older you get, the harder it is for your body to keep you healthy and process nutrients into energy. That's why you feel more sluggish, and your risk of illness or injury increases. It's also why people became interested in NAD supplementation.

The idea is that supplementing with NAD could equate to more energy, better health, and maybe even a longer, healthier life. But before you hit the submit button on your virtual shopping cart, let's uncover what happens to your body when you take NAD supplements.

NAD acts as a coenzyme within the body to help several metabolic processes happen. Essentially, coenzymes assist other molecules in whatever job they're designed to do. Without this metabolic pathway, you would not function.

Then there's DNA, which is easily damaged via life stressors, UV radiation, and alkylating and oxidative molecules. Fortunately, your body is quite good at maintaining DNA integrity by sending in NAD to get the job done. Without getting into the complex nature of DNA repair pathways, NAD calls in molecules to quickly nourish and rejuvenate the damaged DNA cells.

As mentioned earlier, NAD acts as an assistant or middleman, so to speak, making things happen and maximizing communication between cells throughout the body. Without enough NAD, simple processes in the body may take longer, partly because of poor communication. This might look like a compromised immune system (AKA you get sick more often), poor stress response, and an increased risk for obesity, type 2 diabetes , metabolic syndrome, cancer, and other degenerative diseases.

Potential Effects of NAD Supplementation

With the inevitable decline in NAD as you age, your energy production and cellular repair also decline. This reduction leads to energy loss, mental fatigue, wrinkles, and frequent sickness. Though there's no way to stop this from happening, you can fill in the gaps with NAD supplementation.

Boosted Cellular Energy

When you eat food , it's broken down into molecules to produce energy in the form of ATP. During this process, NAD acts as an enzyme facilitating the conversion of NADH to ATP. Without enough NAD, the body has a difficult time producing ATP in a timely manner, which can lead to feeling sluggish and tired. Supplementing with NAD could increase cellular energy production by boosting ATP levels, potentially improving metabolism and reducing fatigue.

Research on people with chronic fatigue syndrome evaluated the safety and effectiveness of NAD supplementation on fatigue, quality of life, and sleep quality. The overall consensus is that supplementing with NAD has a positive effect on quality of life and improves health parameters like anxiety, energy, and immune system response.

Enhanced Cellular Repair

Because NAD is involved in mechanisms of DNA repair that help protect cells from damage caused by stress, oxidative damage, and aging, higher levels (as a result of supplementation) may promote a more efficient process.

For instance, one study investigated the role of NAD supplementation on DNA repair. Researchers found that supplementation decreased the accumulation of endogenous DNA damage and improved DNA repair capacity. Ultimately, supplementing with NAD may positively impact DNA repair in the context of aging and in neurodegenerative diseases like Alzheimer's disease.

Improved Cellular Communication

Cellular communication is essential for overall health, well-being, and longevity. If cells cannot get messages to each other promptly, aging will advance.

Because NAD acts like an assistant, its job is to ensure that things happen quickly and efficiently. Restoring NAD to youthful levels means your body's checking things off its daily to-do list fast. In doing so, it restores cell-signaling function and reduces stress.

Potential Negative Side Effects

If all of this sounds too good to be true, it might be because some studies report mild side effects of using NAD supplements, especially in high doses. Excess of 500 milligrams of nicotinamide isn't recommended. Research shows that excess supplementation increases the risk of vascular disease, neurodegenerative disease, and chronic kidney disease. Other symptoms, including diarrhea, easy bruising, and increased wound bleeding, may be observed. Excess of 3,000 milligrams may cause nausea, vomiting, and liver damage.

It's important to consult with a healthcare provider before taking NAD, especially if you have a preexisting condition or are taking medication.

While there is existing research on NAD supplementation in humans, it is not complete or conclusive. There is always room for ongoing studies and findings.

That said, there are some promising benefits to using a NAD supplement for conditions like age-related cognitive decline, neurodegenerative diseases (Alzheimer's, Parkinson's), and cardiovascular health. Here's what research has shown so far.

  • Age-related cognitive decline: Research on using NAD supplements for cognitive decline is promising. However, the majority of the findings were in animal studies. Despite this, a recent review found that NAD supplementation may preserve cognitive health across a variety of age-related diseases, including Alzheimer's disease and dementia.
  • Neurodegenerative diseases: Healthy brain aging is extremely important especially in the prevention of neurodegenerative diseases like Alzheimer's, Parkinson's, and Huntington diseases. During normal brain aging, NAD levels decline. Over time this leads to cellular damage, improper cellular waste disposal, impaired stress response, DNA repair delay, stem cell exhaustion, and inflammation. Supplementing with NAD may return NAD to healthy levels and prevent further damage to brain cells. Though promising, more research in humans is needed.
  • Cardiovascular health: Cardiovascular diseases often involve metabolic disorders (abnormal lipid and glucose metabolism), oxidative stress (cell damage), and inflammation (poor cell signaling). These stress markers make it difficult for the heart to function properly and contribute to cardiovascular disease. Even within the heart, NAD homeostasis is important. One review found that boosting levels of NAD with supplements has favorable effects on heart health and may have a profound influence on cardiovascular disease.

The two most common delivery methods for NAD are oral tablets and IV injections. NAD tablets are available online, usually as an NAD precursor nicotinamide riboside (NR) and/or niacin (NADH). These precursor supplements strongly influence positive NAD levels.

NAD+ IV injections are available at local clinics and are generally not covered by insurance. These intravenous injections use a more immediate source of NAD. However, it must be dripped over the course of at least two hours to prevent adverse events like rapid heartbeat, nausea, and anxiousness.

The best dosage and delivery method of NAD is still under investigation. However, most reported side effects occur only in doses higher than those used clinically or for those used for daily supplementation. Therefore, take the dosage recommended by a healthcare provider and avoid exceeding this to prevent side effects and potential health risks.

If you're not ready to take the NAD supplement leap, there are ways to boost NAD naturally with simple lifestyle changes. Here's what you need to know.

  • Exercise : Regular aerobic and resistance training exercise has a regenerative effect on NAD levels and can even return them to young adult levels. A 2019 study found that aerobic and resistance training increase NAMPT, the precursor to NAD.
  • Circadian Rhythm & Sleep : Your circadian rhythm (AKA internal 24-hour clock) and sleep patterns directly correlate with NAD+ production. Getting enough sleep is key for maintaining adequate levels—sleep deprivation and disrupted sleep patterns are associated with decreased NAD+ synthesis.
  • Environmental Stress (Heat/Cold) : Cold plunge pools and saunas may help boost NAD. Exposure to extreme hot and cold temperatures may place extra stress on the body, triggering the production of NAD+.
  • Certain foods : A balanced diet can provide all the NAD you need, especially if you include foods rich in vitamin B3 (niacin) and leafy green vegetables. Vitamin B3 (aka niacin) can easily be found in meat, fish, dairy products, avocado, and whole grains and is the precursor to NAD. Niacin converts to NAD for use in cellular metabolism. Tryptophan, a precursor for NAD, is found in turkey, chicken, fish, nuts, and soy. Leafy green vegetables, including broccoli, colored cabbages, and green onions, also contain compounds that positively influence NAD levels.

Bottom Line

While NAD is promising as an anti-aging powerhouse, there are still many unknowns regarding NAD supplementation. Before making any changes to your healthcare routine and supplement lineup, talk to a healthcare provider, such as a registered dietitian to ensure you're making the right choice for you. In the meantime, fueling your body with nourishing foods and living a healthy lifestyle free from tobacco and other stressors is the cornerstone for promoting cellular health and overall well-being.

Covarrubias AJ, Perrone R, Grozio A, Verdin E. NAD+ metabolism and its roles in cellular processes during ageing. Nat Rev Mol Cell Biol . 2021 Feb;22(2):119-141. doi:10.1038/s41580-020-00313-x

Cantó C, Menzies KJ, Auwerx J. NAD(+) Metabolism and the Control of Energy Homeostasis: A Balancing Act between Mitochondria and the Nucleus. Cell Metab . 2015 Jul 7;22(1):31-53. doi:10.1016/j.cmet.2015.05.023

Ruszkiewicz JA, Bürkle A, Mangerich A. Fueling genome maintenance: On the versatile roles of NAD+ in preserving DNA integrity. J Biol Chem . 2022 Jun;298(6):102037. doi:10.1016/j.jbc.2022.102037

Poljšak B, Kovač V, Špalj S, Milisav I. The central role of the NAD+ molecule in the development of aging and the prevention of chronic age-related diseases: Strategies for NAD+ modulation. Int J Mol Sci . 2023 Feb 3;24(3):2959. doi:10.3390/ijms24032959

Mehmel M, Jovanović N, Spitz U. Nicotinamide Riboside-The Current State of Research and Therapeutic Uses . Nutrients . 2020 May 31;12(6):1616. doi:10.3390/nu12061616

Gindri IM, Ferrari G, Pinto LPS, Bicca J, Dos Santos IK, Dallacosta D, Roesler CRM. Evaluation of safety and effectiveness of NAD in different clinical conditions: a systematic review. Am J Physiol Endocrinol Metab . 2024 Apr 1;326(4):E417-E427. doi:10.1152/ajpendo.00242.2023

Wilk A, Hayat F, Cunningham R, Li J, Garavaglia S, Zamani L, Ferraris DM, Sykora P, Andrews J, Clark J, Davis A, Chaloin L, Rizzi M, Migaud M, Sobol RW. Extracellular NAD+ enhances PARP-dependent DNA repair capacity independently of CD73 activity. Sci Rep. 2020 Jan 20;10(1):651. doi:10.1038/s41598-020-57506-9

Poljšak B, Kovač V, Špalj S, Milisav I. The central role of the nad+ molecule in the development of aging and the prevention of chronic age-related diseases: strategies for nad+ modulation .  IJMS . 2023;24(3):2959. doi:10.3390/ijms24032959

National Institutes of Health: Office of Dietary Supplements. Consumer Fact Sheets: Niacin .

Campbell JM. Supplementation with NAD+ and its precursors to prevent cognitive decline across disease contexts. Nutrients. 2022 Aug 7;14(15):3231. doi:10.3390/nu14153231

Lautrup S, Sinclair DA, Mattson MP, Fang EF. NAD+ in brain aging and neurodegenerative disorders . Cell Metab. 2019 Oct 1;30(4):630-655. doi:10.1016/j.cmet.2019.09.001

Zhang X, Zhang Y, Sun A, Ge J. The effects of nicotinamide adenine dinucleotide in cardiovascular diseases: Molecular mechanisms, roles and therapeutic potential . Genes Dis . 2021 Apr 24;9(4):959-972. doi:10.1016/j.gendis.2021.04.001

Okabe K, Yaku K, Uchida Y, Fukamizu Y, Sato T, Sakurai T, Tobe K, Nakagawa T. Oral Administration of Nicotinamide Mononucleotide Is Safe and Efficiently Increases Blood Nicotinamide Adenine Dinucleotide Levels in Healthy Subjects. Front Nutr. 2022 Apr 11;9:868640. doi:10.3389/fnut.2022.868640

Grant R, Berg J, Mestayer R, Braidy N, Bennett J, Broom S, Watson J. A Pilot Study Investigating Changes in the Human Plasma and Urine NAD+ Metabolome During a 6 Hour Intravenous Infusion of NAD . Front Aging Neurosci . 2019 Sep 12;11:257. doi:10.3389/fnagi.2019.00257

de Guia RM, Agerholm M, Nielsen TS, et al. Aerobic and resistance exercise training reverses age‐dependent decline in NAD  +  salvage capacity in human skeletal muscle .  Physiol Rep . 2019;7(12). doi:10.14814/phy2.14139

Poljsak B, Kovač V, Milisav I. Healthy lifestyle recommendations: do the beneficial effects originate from nad+ amount at the cellular level? Myers JN, ed.  Oxidative Medicine and Cellular Longevity . 2020;2020:1-12. doi:10.1155/2020/8819627

Kirkland JB, Meyer-Ficca ML. Niacin . Adv Food Nutr Res . 2018;83:83-149. doi:10.1016/bs.afnr.2017.11.003

Connell N, Grevendonk L, Fealy C, et al. Nad+-precursor supplementation with l-tryptophan, nicotinic acid, and nicotinamide does not affect mitochondrial function or skeletal muscle function in physically compromised older adults .  The Journal of Nutrition . 2021;151(10):2917-2931. doi:10.1093/jn/nxab193

By Shoshana Pritzker RD, CDN, CSSD, CISSN Shoshana Pritzker RD, CDN is a sports and pediatric dietitian, the owner of Nutrition by Shoshana, and is the author of "Carb Cycling for Weight Loss." Shoshana received her B.S in dietetics and nutrition from Florida International University. She's been writing and creating content in the health, nutrition, and fitness space for over 15 years and is regularly featured in Oxygen Magazine,, and more.

  • Open access
  • Published: 27 May 2024

Current status of community resources and priorities for weed genomics research

  • Jacob Montgomery 1 ,
  • Sarah Morran 1 ,
  • Dana R. MacGregor   ORCID: 2 ,
  • J. Scott McElroy   ORCID: 3 ,
  • Paul Neve   ORCID: 4 ,
  • Célia Neto   ORCID: 4 ,
  • Martin M. Vila-Aiub   ORCID: 5 ,
  • Maria Victoria Sandoval 5 ,
  • Analia I. Menéndez   ORCID: 6 ,
  • Julia M. Kreiner   ORCID: 7 ,
  • Longjiang Fan   ORCID: 8 ,
  • Ana L. Caicedo   ORCID: 9 ,
  • Peter J. Maughan 10 ,
  • Bianca Assis Barbosa Martins 11 ,
  • Jagoda Mika 11 ,
  • Alberto Collavo 11 ,
  • Aldo Merotto Jr.   ORCID: 12 ,
  • Nithya K. Subramanian   ORCID: 13 ,
  • Muthukumar V. Bagavathiannan   ORCID: 13 ,
  • Luan Cutti   ORCID: 14 ,
  • Md. Mazharul Islam 15 ,
  • Bikram S. Gill   ORCID: 16 ,
  • Robert Cicchillo 17 ,
  • Roger Gast 17 ,
  • Neeta Soni   ORCID: 17 ,
  • Terry R. Wright   ORCID: 18 ,
  • Gina Zastrow-Hayes 18 ,
  • Gregory May 18 ,
  • Jenna M. Malone   ORCID: 19 ,
  • Deepmala Sehgal   ORCID: 20 ,
  • Shiv Shankhar Kaundun   ORCID: 20 ,
  • Richard P. Dale 20 ,
  • Barend Juan Vorster   ORCID: 21 ,
  • Bodo Peters 11 ,
  • Jens Lerchl   ORCID: 22 ,
  • Patrick J. Tranel   ORCID: 23 ,
  • Roland Beffa   ORCID: 24 ,
  • Alexandre Fournier-Level   ORCID: 25 ,
  • Mithila Jugulam   ORCID: 15 ,
  • Kevin Fengler 18 ,
  • Victor Llaca   ORCID: 18 ,
  • Eric L. Patterson   ORCID: 14 &
  • Todd A. Gaines   ORCID: 1  

Genome Biology volume  25 , Article number:  139 ( 2024 ) Cite this article

680 Accesses

10 Altmetric

Metrics details

Weeds are attractive models for basic and applied research due to their impacts on agricultural systems and capacity to swiftly adapt in response to anthropogenic selection pressures. Currently, a lack of genomic information precludes research to elucidate the genetic basis of rapid adaptation for important traits like herbicide resistance and stress tolerance and the effect of evolutionary mechanisms on wild populations. The International Weed Genomics Consortium is a collaborative group of scientists focused on developing genomic resources to impact research into sustainable, effective weed control methods and to provide insights about stress tolerance and adaptation to assist crop breeding.

Each year globally, agricultural producers and landscape managers spend billions of US dollars [ 1 , 2 ] and countless hours attempting to control weedy plants and reduce their adverse effects. These management methods range from low-tech (e.g., pulling plants from the soil by hand) to extremely high-tech (e.g., computer vision-controlled spraying of herbicides). Regardless of technology level, effective control methods serve as strong selection pressures on weedy plants and often result in rapid evolution of weed populations resistant to such methods [ 3 , 4 , 5 , 6 , 7 ]. Thus, humans and weeds have been locked in an arms race, where humans develop new or improved control methods and weeds adapt and evolve to circumvent such methods.

Applying genomics to weed science offers a unique opportunity to study rapid adaptation, epigenetic responses, and examples of evolutionary rescue of diverse weedy species in the face of widespread and powerful selective pressures. Furthermore, lessons learned from these studies may also help to develop more sustainable control methods and to improve crop breeding efforts in the face of our ever-changing climate. While other research fields have used genetics and genomics to uncover the basis of many biological traits [ 8 , 9 , 10 , 11 ] and to understand how ecological factors affect evolution [ 12 , 13 ], the field of weed science has lagged behind in the development of genomic tools essential for such studies [ 14 ]. As research in human and crop genetics pushes into the era of pangenomics (i.e., multiple chromosome scale genome assemblies for a single species [ 15 , 16 ]), publicly available genomic information is still lacking or severely limited for the majority of weed species. Recent reviews of current weed genomes identified 26 [ 17 ] and 32 weed species with sequenced genomes [ 18 ]—many assembled to a sub-chromosome level.

Here, we summarize the current state of weed genomics, highlighting cases where genomics approaches have successfully provided insights on topics such as population genetic dynamics, genome evolution, and the genetic basis of herbicide resistance, rapid adaptation, and crop dedomestication. These highlighted investigations all relied upon genomic resources that are relatively rare for weedy species. Throughout, we identify additional resources that would advance the field of weed science and enable further progress in weed genomics. We then introduce the International Weed Genomics Consortium (IWGC), an open collaboration among researchers, and describe current efforts to generate these additional resources.

Evolution of weediness: potential research utilizing weed genomics tools

Weeds can evolve from non-weed progenitors through wild colonization, crop de-domestication, or crop-wild hybridization [ 19 ]. Because the time span in which weeds have evolved is necessarily limited by the origins of agriculture, these non-weed relatives often still exist and can be leveraged through population genomic and comparative genomic approaches to identify the adaptive changes that have driven the evolution of weediness. The ability to rapidly adapt, persist, and spread in agroecosystems are defining features of weedy plants, leading many to advocate agricultural weeds as ideal candidates for studying rapid plant adaptation [ 20 , 21 , 22 , 23 ]. The insights gained from applying plant ecological approaches to the study of rapid weed adaptation will move us towards the ultimate goals of mitigating such adaptation and increasing the efficacy of crop breeding and biotechnology [ 14 ].

Biology and ecological genomics of weeds

The impressive community effort to create and maintain resources for Arabidopsis thaliana ecological genomics provides a motivating example for the emerging study of weed genomics [ 24 , 25 , 26 , 27 ]. Arabidopsis thaliana was the first flowering plant species to have its genome fully sequenced [ 28 ] and rapidly became a model organism for plant molecular biology. As weedy genomes become available, collection, maintenance, and resequencing of globally distributed accessions of these species will help to replicate the success found in ecological studies of A. thaliana [ 29 , 30 , 31 , 32 , 33 , 34 , 35 ]. Evaluation of these accessions for traits of interest to produce large phenomics data sets (as in [ 36 , 37 , 38 , 39 , 40 ]) enables genome-wide association studies and population genomics analyses aimed at dissecting the genetic basis of variation in such traits [ 41 ]. Increasingly, these resources (e.g. the 1001 genomes project [ 29 ]) have enabled A. thaliana to be utilized as a model species to explore the eco-evolutionary basis of plant adaptation in a more realistic ecological context. Weedy species should supplement lessons in eco-evolutionary genomics learned from these experiments in A. thaliana .

Untargeted genomic approaches for understanding the evolutionary trajectories of populations and the genetic basis of traits as described above rely on the collection of genotypic information from across the genome of many individuals. While whole-genome resequencing accomplishes this requirement and requires no custom methodology, this approach provides more information than is necessary and is prohibitively expensive in species with large genomes. Development and optimization of genotype-by-sequencing methods for capturing reduced representations of newly sequence genomes like those described by [ 42 , 43 , 44 ] will reduce the cost and computational requirements of genetic mapping and population genetic experiments. Most major weed species do not currently have protocols for stable transformation, a key development in the popularity of A. thaliana as a model organism and a requirement for many functional genomic approaches. Functional validation of genes/variants believed to be responsible for traits of interest in weeds has thus far relied on transiently manipulating endogenous gene expression [ 45 , 46 ] or ectopic expression of a transgene in a model system [ 47 , 48 , 49 ]. While these methods have been successful, few weed species have well-studied viral vectors to adapt for use in virus induced gene silencing. Spray induced gene silencing is another potential option for functional investigation of candidate genes in weeds, but more research is needed to establish reliable delivery and gene knockdown [ 50 ]. Furthermore, traits with complex genetic architecture divergent between the researched and model species may not be amenable to functional genomic approaches using transgenesis techniques in model systems. Developing protocols for reduced representation sequencing, stable transformation, and gene editing/silencing in weeds will allow for more thorough characterization of candidate genetic variants underlying traits of interest.

Beyond rapid adaptation, some weedy species offer an opportunity to better understand co-evolution, like that between plants and pollinators and how their interaction leads to the spread of weedy alleles (Additional File 1 : Table S1). A suite of plant–insect traits has co-evolved to maximize the attraction of the insect pollinator community and the efficiency of pollen deposition between flowers ensuring fruit and seed production in many weeds [ 51 , 52 ]. Genetic mapping experiments have identified genes and genetic variants responsible for many floral traits affecting pollinator interaction including petal color [ 53 , 54 , 55 , 56 ], flower symmetry and size [ 57 , 58 , 59 ], and production of volatile organic compounds [ 60 , 61 , 62 ] and nectar [ 63 , 64 , 65 ]. While these studies reveal candidate genes for selection under co-evolution, herbicide resistance alleles may also have pleiotropic effects on the ecology of weeds [ 66 ], altering plant-pollinator interactions [ 67 ]. Discovery of genes and genetic variants involved in weed-pollinator interaction and their molecular and environmental control may create opportunities for better management of weeds with insect-mediated pollination. For example, if management can disrupt pollinator attraction/interaction with these weeds, the efficiency of reproduction may be reduced.

A more complete understanding of weed ecological genomics will undoubtedly elucidate many unresolved questions regarding the genetic basis of various aspects of weediness. For instance, when comparing populations of a species from agricultural and non-agricultural environments, is there evidence for contemporary evolution of weedy traits selected by agricultural management or were “natural” populations pre-adapted to agroecosystems? Where there is differentiation between weedy and natural populations, which traits are under selection and what is the genetic basis of variation in those traits? When comparing between weedy populations, is there evidence for parallel versus non-parallel evolution of weediness at the phenotypic and genotypic levels? Such studies may uncover fundamental truths about weediness. For example, is there a common phenotypic and/or genotypic basis for aspects of weediness among diverse weed species? The availability of characterized accessions and reference genomes for species of interest are required for such studies but only a few weedy species have these resources developed.

Population genomics

Weed species are certainly fierce competitors, able to outcompete crops and endemic species in their native environment, but they are also remarkable colonizers of perturbed habitats. Weeds achieve this through high fecundity, often producing tens of thousands of seeds per individual plant [ 68 , 69 , 70 ]. These large numbers in terms of demographic population size often combine with outcrossing reproduction to generate high levels of diversity with local effective population sizes in the hundreds of thousands [ 71 , 72 ]. This has two important consequences: weed populations retain standing genetic variation and generate many new mutations, supporting weed success in the face of harsh control. The generation of genomic tools to monitor weed populations at the molecular level is a game-changer to understanding weed dynamics and precisely testing the effect of artificial selection (i.e., management) and other evolutionary mechanisms on the genetic make-up of populations.

Population genomic data, without any environmental or phenotypic information, can be used to scan the genomes of weed and non-weed relatives to identify selective sweeps, pointing at loci supporting weed adaptation on micro- or macro-evolutionary scales. Two recent within-species examples include weedy rice, where population differentiation between weedy and domesticated populations was used to identify the genetic basis of weedy de-domestication [ 73 ], and common waterhemp, where consistent allelic differences among natural and agricultural collections resolved a complex set of agriculturally adaptive alleles [ 74 , 75 ]. A recent comparative population genomic study of weedy barnyardgrass and crop millet species has demonstrated how inter-specific investigations can resolve the signatures of crop and weed evolution [ 76 ] (also see [ 77 ] for a non-weed climate adaptation example). Multiple sequence alignments across numerous species provide complementary insight into adaptive convergence over deeper timescales, even with just one genomic sample per species (e.g., [ 78 , 79 ]). Thus, newly sequenced weed genomes combined with genomes available for closely related crops (outlined by [ 14 , 80 ]) and an effort to identify other non-weed wild relatives will be invaluable in characterizing the genetic architecture of weed adaptation and evolution across diverse species.

Weeds experience high levels of genetic selection, both artificial in response to agricultural practices and particularly herbicides, and natural in response to the environmental conditions they encounter [ 81 , 82 ]. Using genomic analysis to identify loci that are the targets of selection, whether natural or artificial, would point at vulnerabilities that could be leveraged against weeds to develop new and more sustainable management strategies [ 83 ]. This is a key motivation to develop genotype-by-environment association (GEA) and selective sweep scan approaches, which allow researchers to resolve the molecular basis of multi-dimensional adaptation [ 84 , 85 ]. GEA approaches, in particular, have been widely used on landscape-wide resequencing collections to determine the genetic basis of climate adaptation (e.g., [ 27 , 86 , 87 ]), but have yet to be fully exploited to diagnose the genetic basis of the various aspects of weediness [ 88 ]. Armed with data on environmental dimensions of agricultural settings, such as focal crop, soil quality, herbicide use, and climate, GEA approaches can help disentangle how discrete farming practices have influenced the evolution of weediness and resolve broader patterns of local adaptation across a weed’s range. Although non-weedy relatives are not technically required for GEA analyses, inclusion of environmental and genomic data from weed progenitors can further distinguish genetic variants underpinning weed origins from those involved in local adaptation.

New weeds emerge frequently [ 89 ], either through hybridization between species as documented for sea beet ( Beta vulgaris ssp. maritima) hybridizing with crop beet to produce progeny that are well adapted to agricultural conditions [ 90 , 91 , 92 ], or through the invasion of alien species that find a new range to colonize. Biosecurity measures are often in place to stop the introduction of new weeds; however, the vast scale of global agricultural commodity trade precludes the possibility of total control. Population genomic analysis is now able to measure gene flow between populations [ 74 , 93 , 94 , 95 ] and identify populations of origin for invasive species including weeds [ 96 , 97 , 98 ]. For example, the invasion route of the pest fruitfly Drosophila suzukii from Eastern Asia to North America and Europe through Hawaii was deciphered using Approximate Bayesian Computation on high-throughput sequencing data from a global sample of multiple populations [ 99 ]. Genomics can also be leveraged to predict invasion rather than explain it. The resequencing of a global sample of common ragweed ( Ambrosia artemisiifolia L.) elucidated a complex invasion route whereby Europe was invaded by multiple introductions of American ragweed that hybridized in Europe prior to a subsequent introduction to Australia [ 100 , 101 ]. In this context, the use of genomically informed species distribution models helps assess the risk associated with different source populations, which in the case of common ragweed, suggests that a source population from Florida would allow ragweed to invade most of northern Australia [ 102 ]. Globally coordinated research efforts to understand potential distribution models could support the transformation of biosecurity from perspective analysis towards predictive risk assessment.

Herbicide resistance and weed management

Herbicide resistance is among the numerous weedy traits that can evolve in plant populations exposed to agricultural selection pressures. Over-reliance on herbicides to control weeds, along with low diversity and lack of redundancy in weed management strategies, has resulted in globally widespread herbicide resistance [ 103 ]. To date, 272 herbicide-resistant weed species have been reported worldwide, and at least one resistance case exists for 21 of the 31 existing herbicide sites of action [ 104 ]—significantly limiting chemical weed control options available to agriculturalists. This limitation of control options is exacerbated by the recent lack of discovery of herbicides with new sites of action [ 105 ].

Herbicide resistance may result from several different physiological mechanisms. Such mechanisms have been classified into two main groups, target-site resistance (TSR) [ 4 , 106 ] and non-target-site resistance (NTSR) [ 4 , 107 ]. The first group encompasses changes that reduce binding affinity between a herbicide and its target [ 108 ]. These changes may provide resistance to multiple herbicides that have a common biochemical target [ 109 ] and can be effectively managed through mixture and/or rotation of herbicides targeting different sites of action [ 110 ]. The second group (NTSR), includes alterations in herbicide absorption, translocation, sequestration, and/or metabolism that may lead to unpredictable pleotropic cross-resistance profiles where structurally and functionally diverse herbicides are rendered ineffective by one or more genetic variant(s) [ 47 ]. This mechanism of resistance threatens not only the efficacy of existing herbicidal chemistries, but also ones yet to be discovered. While TSR is well understood because of the ease of identification and molecular characterization of target site variants, NTSR mechanisms are significantly more challenging to research because they are often polygenic, and the resistance causing element(s) are not well understood [ 111 ].

Improving the current understanding of metabolic NTSR mechanisms is not an easy task, since genes of diverse biochemical functions are involved, many of which exist as extensive gene families [ 109 , 112 ]. Expression changes of NTSR genes have been implicated in several resistance cases where the protein products of the genes are functionally equivalent across sensitive and resistant plants, but their relative abundance leads to resistance. Thus, regulatory elements of NTSR genes have been scrutinized to understand their role in NTSR mechanisms [ 113 ]. Similarly, epigenetic modifications have been hypothesized to play a role in NTSR, with much remaining to be explored [ 114 , 115 , 116 ]. Untargeted approaches such as genome-wide association, selective sweep scans, linkage mapping, RNA-sequencing, and metabolomic profiling have proven helpful to complement more specific biochemical- and chemo-characterization studies towards the elucidation of NTSR mechanisms as well as their regulation and evolution [ 47 , 117 , 118 , 119 , 120 , 121 , 122 , 123 , 124 ]. Even in cases where resistance has been attributed to TSR, genetic mapping approaches can detect other NTSR loci contributing to resistance (as shown by [ 123 ]) and provide further evidence for the role of TSR mutations across populations. Knowledge of the genetic basis of NTSR will aid the rational design of herbicides by screening new compounds for interaction with newly discovered NTSR proteins during early research phases and by identifying conserved chemical structures that interact with these proteins that should be avoided in small molecule design.

Genomic resources can also be used to predict the protein structure for novel herbicide target site and metabolism genes. This will allow for prediction of efficacy and selectivity for new candidate herbicides in silico to increase herbicide discovery throughput as well as aid in the design and development of next-generation technologies for sustainable weed management. Proteolysis targeting chimeras (PROTACs) have the potential to bind desired targets with great selectivity and degrade proteins by utilizing natural protein ubiquitination and degradation pathways within plants [ 125 ]. Spray-induced gene silencing in weeds using oligonucleotides has potential as a new, innovative, and sustainable method for weed management, but improved methods for design and delivery of oligonucleotides are needed to make this technique a viable management option [ 50 ]. Additionally, success in the field of pharmaceutical drug discovery in the development of molecules modulating protein–protein interactions offers another potential avenue towards the development of herbicides with novel targets [ 126 , 127 ]. High-quality reference genomes allow for the design of new weed management technologies like the ones listed here that are specific to—and effective across—weed species but have a null effect on non-target organisms.

Comparative genomics and genome biology

The genomes of weed species are as diverse as weed species themselves. Weeds are found across highly diverged plant families and often have no phylogenetically close model or crop species relatives for comparison. On all measurable metrics, weed genomes run the gamut. Some have smaller genomes like Cyperus spp. (~ 0.26 Gb) while others are larger, such as Avena fatua (~ 11.1 Gb) (Table  1 ). Some have high heterozygosity in terms of single-nucleotide polymorphisms, such as the Amaranthus spp., while others are primarily self-pollinated and quite homozygous, such as Poa annua [ 128 , 129 ]. Some are diploid such as Conyza canadensis and Echinochloa haploclada while others are polyploid such as C. sumetrensis , E. crus-galli , and E. colona [ 76 ]. The availability of genomic resources in these diverse, unexplored branches of the tree of life allows us to identify consistencies and anomalies in the field of genome biology.

The weed genomes published so far have focused mainly on weeds of agronomic crops, and studies have revolved around their ability to resist key herbicides. For example, genomic resources were vital in the elucidation of herbicide resistance cases involving target site gene copy number variants (CNVs). Gene CNVs of 5-enolpyruvylshikimate-3-phosphate synthase ( EPSPS ) have been found to confer resistance to the herbicide glyphosate in diverse weed species. To date, nine species have independently evolved EPSPS CNVs, and species achieve increased EPSPS copy number via different mechanisms [ 153 ]. For instance, the EPSPS CNV in Bassia scoparia is caused by tandem duplication, which is accredited to transposable element insertions flanking EPSPS and subsequent unequal crossing over events [ 154 , 155 ]. In Eleusine indica , a EPSPS CNV was caused by translocation of the EPSPS locus into the subtelomere followed by telomeric sequence exchange [ 156 ]. One of the most fascinating genome biology discoveries in weed science has been that of extra-chromosomal circular DNAs (eccDNAs) that harbor the EPSPS gene in the weed species Amaranthus palmeri [ 157 , 158 ]. In this case, the eccDNAs autonomously replicate separately from the nuclear genome and do not reintegrate into chromosomes, which has implications for inheritance, fitness, and genome structure [ 159 ]. These discoveries would not have been possible without reference assemblies of weed genomes, next-generation sequencing, and collaboration with experts in plant genomics and bioinformatics.

Another question that is often explored with weedy genomes is the nature and composition of gene families that are associated with NTSR. Gene families under consideration often include cytochrome P450s (CYPs), glutathione- S -transferases (GSTs), ABC transporters, etc. Some questions commonly considered with new weed genomes include how many genes are in each of these gene families, where are they located, and which weed accessions and species have an over-abundance of them that might explain their ability to evolve resistance so rapidly [ 76 , 146 , 160 , 161 ]? Weed genome resources are necessary to answer questions about gene family expansion or contraction during the evolution of weediness, including the role of polyploidy in NTSR gene family expansion as explored by [ 162 ].

Translational research and communication with weed management stakeholders

Whereas genomics of model plants is typically aimed at addressing fundamental questions in plant biology, and genomics of crop species has the obvious goal of crop improvement, goals of genomics of weedy plants also include the development of more effective and sustainable strategies for their management. Weed genomic resources assist with these objectives by providing novel molecular ecological and evolutionary insights from the context of intensive anthropogenic management (which is lacking in model plants), and offer knowledge and resources for trait discovery for crop improvement, especially given that many wild crop relatives are also important agronomic weeds (e.g., [ 163 ]). For instance, crop-wild relatives are valuable for improving crop breeding for marginal environments [ 164 ]. Thus, weed genomics presents unique opportunities and challenges relative to plant genomics more broadly. It should also be noted that although weed science at its core is an applied discipline, it draws broadly from many scientific disciplines such as, plant physiology, chemistry, ecology, and evolutionary biology, to name a few. The successful integration of weed-management strategies, therefore, requires extensive collaboration among individuals collectively possessing the necessary expertise [ 165 ].

With the growing complexity of herbicide resistance management, practitioners are beginning to recognize the importance of understanding resistance mechanisms to inform appropriate management tactics [ 14 ]. Although weed science practitioners do not need to understand the technical details of weed genomics, their appreciation of the power of weed genomics—together with their unique insights from field observations—will yield novel opportunities for applications of weed genomics to weed management. In particular, combining field management history with information on weed resistance mechanisms is expected to provide novel insights into evolutionary trajectories (e.g. [ 6 , 166 ]), which can be utilized for disrupting evolutionary adaptation. It can be difficult to obtain field history information from practitioners, but developing an understanding among them of the importance of such information can be invaluable.

Development of weed genomics resources by the IWGC

Weed genomics is a fast-growing field of research with many recent breakthroughs and many unexplored areas of study. The International Weed Genomics Consortium (IWGC) started in 2021 to address the roadblocks listed above and to promote the study of weedy plants. The IWGC is an open collaboration among academic, government, and industry researchers focused on producing genomic tools for weedy species from around the world. Through this collaboration, our initial aim is to provide chromosome-level reference genome assemblies for at least 50 important weedy species from across the globe that are chosen based on member input, economic impact, and global prevalence (Fig.  1 ). Each genome will include annotation of gene models and repetitive elements and will be freely available through public databases with no intellectual property restrictions. Additionally, future funding of the IWGC will focus on improving gene annotations and supplementing these reference genomes with tools that increase their utility.

figure 1

The International Weed Genomics Consortium (IWGC) collected input from the weed genomics community to develop plans for weed genome sequencing, annotation, user-friendly genome analysis tools, and community engagement

Reference genomes and data analysis tools

The first objective of the IWGC is to provide high-quality genomic resources for agriculturally important weeds. The IWGC therefore created two main resources for information about, access to, or analysis of weed genomic data (Fig.  1 ). The IWGC website (available at [ 167 ]) communicates the status and results of genome sequencing projects, information on training and funding opportunities, upcoming events, and news in weed genomics. It also contains details of all sequenced species including genome size, ploidy, chromosome number, herbicide resistance status, and reference genome assembly statistics. The IWGC either compiles existing data on genome size, ploidy, and chromosome number, or obtains the data using flow cytometry and cytogenetics (Fig.  1 ; Additional File 2 : Fig S1-S4). Through this website, users can request an account to access our second main resource, an online genome database called WeedPedia (accessible at [ 168 ]), with an account that is created within 3–5 working days of an account request submission. WeedPedia hosts IWGC-generated and other relevant publicly accessible genomic data as well as a suite of bioinformatic tools. Unlike what is available for other fields, weed science did not have a centralized hub for genomics information, data, and analysis prior to the IWGC. Our intention in creating WeedPedia is to encourage collaboration and equity of access to information across the research community. Importantly, all genome assemblies and annotations from the IWGC (Table  1 ), along with the raw data used to produce them, will be made available through NCBI GenBank. Upon completion of a 1-year sponsoring member data confidentiality period for each species (dates listed in Table  1 ), scientific teams within the IWGC produce the first genome-wide investigation to submit for publication including whole genome level analyses on genes, gene families, and repetitive sequences as well as comparative analysis with other species. Genome assemblies and data will be publicly available through NCBI as part of these initial publications for each species.

WeedPedia is a cloud-based omics database management platform built from the software “CropPedia” and licensed from KeyGene (Wageningen, The Netherlands). The interface allows users to access, visualize, and download genome assemblies along with structural and functional annotation. The platform includes a genome browser, comparative map viewer, pangenome tools, RNA-sequencing data visualization tools, genetic mapping and marker analysis tools, and alignment capabilities that allow searches by keyword or sequence. Additionally, genes encoding known target sites of herbicides have been specially annotated, allowing users to quickly identify and compare these genes of interest. The platform is flexible, making it compatible with future integration of other data types such as epigenetic or proteomic information. As an online platform with a graphical user interface, WeedPedia provides user-friendly, intuitive tools that encourage users to integrate genomics into their research while also allowing more advanced users to download genomic data to be used in custom analysis pipelines. We aspire for WeedPedia to mimic the success of other public genomic databases such as NCBI, CoGe, Phytozome, InsectBase, and Mycocosm to name a few. WeedPedia currently hosts reference genomes for 40 species (some of which are currently in their 1-year confidentiality period) with additional genomes in the pipeline to reach a currently planned total of 55 species (Table  1 ). These genomes include both de novo reference genomes generated or in progress by the IWGC (31 species; Table  1 ), and publicly available genome assemblies of 24 weedy or related species that were generated by independent research groups (Table  2 ). As of May 2024, WeedPedia has over 370 registered users from more than 27 countries spread across 6 continents.

The IWGC reference genomes are generated in partnership with the Corteva Agriscience Genome Center of Excellence (Johnston, Iowa) using a combination of single-molecule long-read sequencing, optical genome maps, and chromosome conformation mapping. This strategy has already yielded highly contiguous, phased, chromosome-level assemblies for 26 weed species, with additional assemblies currently in progress (Table  1 ). The IWGC assemblies have been completed as single or haplotype-resolved double-haplotype pseudomolecules in inbreeding and outbreeding species, respectively, with multiple genomes being near gapless. For example, the de novo assemblies of the allohexaploids Conyza sumatrensis and Chenopodium album have all chromosomes captured in single scaffolds and most chromosomes being gapless from telomere to telomere. Complementary full-length isoform (IsoSeq) sequencing of RNA collected from diverse tissue types and developmental stages assists in the development of gene models during annotation.

As with accessibility of data, a core objective of the IWGC is to facilitate open access to sequenced germplasm when possible for featured species. Historically, the weed science community has rarely shared or adopted standard germplasm (e.g., specific weed accessions). The IWGC has selected a specific accession of each species for reference genome assembly (typically susceptible to herbicides). In collaboration with a parallel effort by the Herbicide Resistant Plants committee of the Weed Science Society of America, seeds of the sequenced weed accessions will be deposited in the United States Department of Agriculture Germplasm Resources Information Network [ 186 ] for broad access by the scientific community and their accession numbers will be listed on the IWGC website. In some cases, it is not possible to generate enough seed to deposit into a public repository (e.g., plants that typically reproduce vegetatively, that are self-incompatible, or that produce very few seeds from a single individual). In these cases, the location of collection for sequenced accessions will at least inform the community where the sequenced individual came from and where they may expect to collect individuals with similar genotypes. The IWGC ensures that sequenced accessions are collected and documented to comply with the Nagoya Protocol on access to genetic resources and the fair and equitable sharing of benefits arising from their utilization under the Convention on Biological Diversity and related Access and Benefit Sharing Legislation [ 187 ]. As additional accessions of weed species are sequenced (e.g., pangenomes are obtained), the IWGC will facilitate germplasm sharing protocols to support collaboration. Further, to simplify the investigation of herbicide resistance, the IWGC will link WeedPedia with the International Herbicide-Resistant Weed Database [ 104 ], an already widely known and utilized database for weed scientists.

Training and collaboration in weed genomics

Beyond producing genomic tools and resources, a priority of the IWGC is to enable the utilization of these resources across a wide range of stakeholders. A holistic approach to training is required for weed science generally [ 188 ], and we would argue even more so for weed genomics. To accomplish our training goals, the IWGC is developing and delivering programs aimed at the full range of IWGC stakeholders and covering a breadth of relevant topics. We have taken care to ensure our approaches are diverse as to provide training to researchers with all levels of existing experience and differing reasons for engaging with these tools. Throughout, the focus is on ensuring that our training and outreach result in impacts that benefit a wide range of stakeholders.

Although recently developed tools are incredibly enabling and have great potential to replace antiquated methodology [ 189 ] and to solve pressing weed science problems [ 14 ], specialized computational skills are required to fully explore and unlock meaning from these highly complex datasets. Collaboration with, or training of, computational biologists equipped with these skills and resources developed by the IWGC will enable weed scientists to expand research programs and better understand the genetic underpinnings of weed evolution and herbicide resistance. To fill existing skill gaps, the IWGC is developing summer bootcamps and online modules directed specifically at weed scientists that will provide training on computational skills (Fig.  1 ). Because successful utilization of the IWGC resources requires more than general computational skills, we have created three targeted workshops that teach practical skills related to genomics databases, molecular biology, and population genomics (available at [ 190 ]). The IWGC has also hosted two official conference meetings, one in September of 2021 and one in January of 2023, with more conferences planned. These conferences have included invited speakers to present successful implementations of weed genomics, educational workshops to build computational skills, and networking opportunities for research to connect and collaborate.

Engagement opportunities during undergraduate degrees have been shown to improve academic outcomes [ 191 , 192 ]. As one activity to help achieve this goal, the IWGC has sponsored opportunities for US undergraduates to undertake a 10-week research experience, which includes an introduction to bioinformatics, a plant genomics research project that results in a presentation, and access to career building opportunities in diverse workplace environments. To increase equitable access to conferences and professional communities, we supported early career researchers to attend the first two IWGC conferences in the USA as well as workshops and bootcamps in Europe, South America, and Australia. These hybrid or in-person travel grants are intentionally designed to remove barriers and increase participation of individuals from backgrounds and experiences currently underrepresented within weed/plant science or genomics [ 193 ]. Recipients of these travel awards gave presentations and gained the measurable benefits that come from either virtual or in-person participation in conferences [ 194 ]. Moving forward, weed scientists must amass skills associated with genomic analyses and collaborate with other area experts to fully leverage resources developed by the IWGC.

The tools generated through the IWGC will enable many new research projects with diverse objectives like those listed above. In summary, contiguous genome assemblies and complete annotation information will allow weed scientists to join plant breeders in the use of genetic mapping for many traits including stress tolerance, plant architecture, and herbicide resistance (especially important for cases of NTSR). These assemblies will also allow for investigations of population structure, gene flow, and responses to evolutionary mechanisms like genetic bottlenecking and artificial selection. Understanding gene sequences across diverse weed species will be vital in modeling new herbicide target site proteins and designing novel effective herbicides with minimal off-target effects. The IWGC website will improve accessibility to weed genomics data by providing a single hub for reference genomes as well as phenotypic and genotypic information for accessions shared with the IWGC. Deposition of sequenced germplasm into public repositories will ensure that researchers are able to access and utilize these accessions in their own research to make the field more standardized and equitable. WeedPedia allows users of all backgrounds to quickly access information of interest such as herbicide target site gene sequence or subcellular localization of protein products for different genes. Users can also utilize server-based tools such as BLAST and genome browsing similar to other public genomic databases. Finally, the IWGC is committed to training and connecting weed genomicists through hosting trainings, workshops, and conferences.


Weeds are unique and fascinating plants, having significant impacts on agriculture and ecosystems; and yet, aspects of their biology, ecology, and genetics remain poorly understood. Weeds represent a unique area within plant biology, given their repeated rapid adaptation to sudden and severe shifts in the selective landscape of anthropogenic management practices. The production of a public genomics database with reference genomes and annotations for over 50 weed species represents a substantial step forward towards research goals that improve our understanding of the biology and evolution of weeds. Future work is needed to improve annotations, particularly for complex gene families involved in herbicide detoxification, structural variants, and mobile genetic elements. As reference genome assemblies become available; standard, affordable methods for gathering genotype information will allow for the identification of genetic variants underlying traits of interest. Further, methods for functional validation and hypothesis testing are needed in weeds to validate the effect of genetic variants detected through such experiments, including systems for transformation, gene editing, and transient gene silencing and expression. Future research should focus on utilizing weed genomes to investigate questions about evolutionary biology, ecology, genetics of weedy traits, and weed population dynamics. The IWGC plans to continue the public–private partnership model to host the WeedPedia database over time, integrate new datasets such as genome resequencing and transcriptomes, conduct trainings, and serve as a research coordination network to ensure that advances in weed science from around the world are shared across the research community (Fig.  1 ). Bridging basic plant genomics with translational applications in weeds is needed to deliver on the potential of weed genomics to improve weed management and crop breeding.

Availability of data and materials

All genome assemblies and related sequencing data produced by the IWGC will be available through NCBI as part of publications reporting the first genome-wide analysis for each species.

Gianessi LP, Nathan PR. The value of herbicides in U.S. crop production. Weed Technol. 2007;21(2):559–66.

Article   Google Scholar  

Pimentel D, Lach L, Zuniga R, Morrison D. Environmental and economic costs of nonindigenous species in the United States. Bioscience. 2000;50(1):53–65.

Barrett SH. Crop mimicry in weeds. Econ Bot. 1983;37(3):255–82.

Powles SB, Yu Q. Evolution in action: plants resistant to herbicides. Annu Rev Plant Biol. 2010;61:317–47.

Article   CAS   PubMed   Google Scholar  

Thurber CS, Reagon M, Gross BL, Olsen KM, Jia Y, Caicedo AL. Molecular evolution of shattering loci in U.S. weedy rice. Mol Ecol. 2010;19(16):3271–84.

Article   PubMed   PubMed Central   Google Scholar  

Comont D, Lowe C, Hull R, Crook L, Hicks HL, Onkokesung N, et al. Evolution of generalist resistance to herbicide mixtures reveals a trade-off in resistance management. Nat Commun. 2020;11(1):3086.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Ashworth MB, Walsh MJ, Flower KC, Vila-Aiub MM, Powles SB. Directional selection for flowering time leads to adaptive evolution in Raphanus raphanistrum (wild radish). Evol Appl. 2016;9(4):619–29.

Chan EK, Rowe HC, Kliebenstein DJ. Understanding the evolution of defense metabolites in Arabidopsis thaliana using genome-wide association mapping. Genetics. 2010;185(3):991–1007.

Frayling TM, Timpson NJ, Weedon MN, Zeggini E, Freathy RM, Lindgren CM, et al. A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity. Science. 2007;316(5826):889–94.

Harkess A, Zhou J, Xu C, Bowers JE, Van der Hulst R, Ayyampalayam S, et al. The asparagus genome sheds light on the origin and evolution of a young Y chromosome. Nat Commun. 2017;8(1):1279.

Periyannan S, Moore J, Ayliffe M, Bansal U, Wang X, Huang L, et al. The gene Sr33 , an ortholog of barley Mla genes, encodes resistance to wheat stem rust race Ug99. Science. 2013;341(6147):786–8.

Ågren J, Oakley CG, McKay JK, Lovell JT, Schemske DW. Genetic mapping of adaptation reveals fitness tradeoffs in Arabidopsis thaliana . Proc Natl Acad Sci U S A. 2013;110(52):21077–82.

Article   PubMed Central   Google Scholar  

Schartl M, Walter RB, Shen Y, Garcia T, Catchen J, Amores A, et al. The genome of the platyfish, Xiphophorus maculatus , provides insights into evolutionary adaptation and several complex traits. Nat Genet. 2013;45(5):567–72.

Ravet K, Patterson EL, Krähmer H, Hamouzová K, Fan L, Jasieniuk M, et al. The power and potential of genomics in weed biology and management. Pest Manag Sci. 2018;74(10):2216–25.

Hufford MB, Seetharam AS, Woodhouse MR, Chougule KM, Ou S, Liu J, et al. De novo assembly, annotation, and comparative analysis of 26 diverse maize genomes. Science. 2021;373(6555):655–62.

Liao W-W, Asri M, Ebler J, Doerr D, Haukness M, Hickey G, et al. A draft human pangenome reference. Nature. 2023;617(7960):312–24.

Huang Y, Wu D, Huang Z, Li X, Merotto A, Bai L, et al. Weed genomics: yielding insights into the genetics of weedy traits for crop improvement. aBIOTECH. 2023;4:20–30.

Chen K, Yang H, Wu D, Peng Y, Lian L, Bai L, et al. Weed biology and management in the multi-omics era: progress and perspectives. Plant Commun. 2024;5(4):100816.

De Wet JMJ, Harlan JR. Weeds and domesticates: evolution in the man-made habitat. Econ Bot. 1975;29(2):99–108.

Mahaut L, Cheptou PO, Fried G, Munoz F, Storkey J, Vasseur F, et al. Weeds: against the rules? Trends Plant Sci. 2020;25(11):1107–16.

Neve P, Vila-Aiub M, Roux F. Evolutionary-thinking in agricultural weed management. New Phytol. 2009;184(4):783–93.

Article   PubMed   Google Scholar  

Sharma G, Barney JN, Westwood JH, Haak DC. Into the weeds: new insights in plant stress. Trends Plant Sci. 2021;26(10):1050–60.

Vigueira CC, Olsen KM, Caicedo AL. The red queen in the corn: agricultural weeds as models of rapid adaptive evolution. Heredity (Edinb). 2013;110(4):303–11.

Donohue K, Dorn L, Griffith C, Kim E, Aguilera A, Polisetty CR, et al. Niche construction through germination cueing: life-history responses to timing of germination in Arabidopsis thaliana . Evolution. 2005;59(4):771–85.

PubMed   Google Scholar  

Exposito-Alonso M. Seasonal timing adaptation across the geographic range of Arabidopsis thaliana . Proc Natl Acad Sci U S A. 2020;117(18):9665–7.

Fournier-Level A, Korte A, Cooper MD, Nordborg M, Schmitt J, Wilczek AM. A map of local adaptation in Arabidopsis thaliana . Science. 2011;334(6052):86–9.

Hancock AM, Brachi B, Faure N, Horton MW, Jarymowycz LB, Sperone FG, et al. Adaptation to climate across the Arabidopsis thaliana genome. Science. 2011;334(6052):83–6.

Initiative TAG. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana . Nature. 2000;408(6814):796–815.

Alonso-Blanco C, Andrade J, Becker C, Bemm F, Bergelson J, Borgwardt KM, et al. 1,135 genomes reveal the global pattern of polymorphism in Arabidopsis thaliana . Cell. 2016;166(2):481–91.

Durvasula A, Fulgione A, Gutaker RM, Alacakaptan SI, Flood PJ, Neto C, et al. African genomes illuminate the early history and transition to selfing in Arabidopsis thaliana . Proc Natl Acad Sci U S A. 2017;114(20):5213–8.

Frachon L, Mayjonade B, Bartoli C, Hautekèete N-C, Roux F. Adaptation to plant communities across the genome of Arabidopsis thaliana . Mol Biol Evol. 2019;36(7):1442–56.

Fulgione A, Koornneef M, Roux F, Hermisson J, Hancock AM. Madeiran Arabidopsis thaliana reveals ancient long-range colonization and clarifies demography in Eurasia. Mol Biol Evol. 2018;35(3):564–74.

Fulgione A, Neto C, Elfarargi AF, Tergemina E, Ansari S, Göktay M, et al. Parallel reduction in flowering time from de novo mutations enable evolutionary rescue in colonizing lineages. Nat Commun. 2022;13(1):1461.

Kasulin L, Rowan BA, León RJC, Schuenemann VJ, Weigel D, Botto JF. A single haplotype hyposensitive to light and requiring strong vernalization dominates Arabidopsis thaliana populations in Patagonia. Argentina Mol Ecol. 2017;26(13):3389–404.

Picó FX, Méndez-Vigo B, Martínez-Zapater JM, Alonso-Blanco C. Natural genetic variation of Arabidopsis thaliana is geographically structured in the Iberian peninsula. Genetics. 2008;180(2):1009–21.

Atwell S, Huang YS, Vilhjálmsson BJ, Willems G, Horton M, Li Y, et al. Genome-wide association study of 107 phenotypes in Arabidopsis thaliana inbred lines. Nature. 2010;465(7298):627–31.

Flood PJ, Kruijer W, Schnabel SK, van der Schoor R, Jalink H, Snel JFH, et al. Phenomics for photosynthesis, growth and reflectance in Arabidopsis thaliana reveals circadian and long-term fluctuations in heritability. Plant Methods. 2016;12(1):14.

Marchadier E, Hanemian M, Tisné S, Bach L, Bazakos C, Gilbault E, et al. The complex genetic architecture of shoot growth natural variation in Arabidopsis thaliana . PLoS Genet. 2019;15(4):e1007954.

Tisné S, Serrand Y, Bach L, Gilbault E, Ben Ameur R, Balasse H, et al. Phenoscope: an automated large-scale phenotyping platform offering high spatial homogeneity. Plant J. 2013;74(3):534–44.

Tschiersch H, Junker A, Meyer RC, Altmann T. Establishment of integrated protocols for automated high throughput kinetic chlorophyll fluorescence analyses. Plant Methods. 2017;13:54.

Chen X, MacGregor DR, Stefanato FL, Zhang N, Barros-Galvão T, Penfield S. A VEL3 histone deacetylase complex establishes a maternal epigenetic state controlling progeny seed dormancy. Nat Commun. 2023;14(1):2220.

Choi M, Scholl UI, Ji W, Liu T, Tikhonova IR, Zumbo P, et al. Genetic diagnosis by whole exome capture and massively parallel DNA sequencing. Proc Natl Acad Sci U S A. 2009;106(45):19096–101.

Davey JW, Blaxter ML. RADSeq: next-generation population genetics. Brief Funct Genomics. 2010;9(5–6):416–23.

Elshire RJ, Glaubitz JC, Sun Q, Poland JA, Kawamoto K, Buckler ES, et al. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS ONE. 2011;6(5):e19379.

MacGregor DR. What makes a weed a weed? How virus-mediated reverse genetics can help to explore the genetics of weediness. Outlooks Pest Manag. 2020;31(5):224–9.

Mellado-Sánchez M, McDiarmid F, Cardoso V, Kanyuka K, MacGregor DR. Virus-mediated transient expression techniques enable gene function studies in blackgrass. Plant Physiol. 2020;183(2):455–9.

Dimaano NG, Yamaguchi T, Fukunishi K, Tominaga T, Iwakami S. Functional characterization of Cytochrome P450 CYP81A subfamily to disclose the pattern of cross-resistance in Echinochloa phyllopogon . Plant Mol Biol. 2020;102(4–5):403–16.

de Figueiredo MRA, Küpper A, Malone JM, Petrovic T, de Figueiredo ABTB, Campagnola G, et al. An in-frame deletion mutation in the degron tail of auxin coreceptor IAA2 confers resistance to the herbicide 2,4-D in Sisymbrium orientale . Proc Natl Acad Sci U S A. 2022;119(9):e2105819119.

Patzoldt WL, Hager AG, McCormick JS, Tranel PJ. A codon deletion confers resistance to herbicides inhibiting protoporphyrinogen oxidase. Proc Natl Acad Sci U S A. 2006;103(33):12329–34.

Zabala-Pardo D, Gaines T, Lamego FP, Avila LA. RNAi as a tool for weed management: challenges and opportunities. Adv Weed Sci. 2022;40(spe1):e020220096.

Fattorini R, Glover BJ. Molecular mechanisms of pollination biology. Annu Rev Plant Biol. 2020;71:487–515.

Rollin O, Benelli G, Benvenuti S, Decourtye A, Wratten SD, Canale A, et al. Weed-insect pollinator networks as bio-indicators of ecological sustainability in agriculture. A review Agron Sustain Dev. 2016;36(1):8.

Irwin RE, Strauss SY. Flower color microevolution in wild radish: evolutionary response to pollinator-mediated selection. Am Nat. 2005;165(2):225–37.

Ma B, Wu J, Shi T-L, Yang Y-Y, Wang W-B, Zheng Y, et al. Lilac ( Syringa oblata ) genome provides insights into its evolution and molecular mechanism of petal color change. Commun Biol. 2022;5(1):686.

Xing A, Wang X, Nazir MF, Zhang X, Wang X, Yang R, et al. Transcriptomic and metabolomic profiling of flavonoid biosynthesis provides novel insights into petals coloration in Asian cotton ( Gossypium arboreum L.). BMC Plant Biol. 2022;22(1):416.

Zheng Y, Chen Y, Liu Z, Wu H, Jiao F, Xin H, et al. Important roles of key genes and transcription factors in flower color differences of Nicotiana alata . Genes (Basel). 2021;12(12):1976.

Krizek BA, Anderson JT. Control of flower size. J Exp Bot. 2013;64(6):1427–37.

Powell AE, Lenhard M. Control of organ size in plants. Curr Biol. 2012;22(9):R360–7.

Spencer V, Kim M. Re"CYC"ling molecular regulators in the evolution and development of flower symmetry. Semin Cell Dev Biol. 2018;79:16–26.

Amrad A, Moser M, Mandel T, de Vries M, Schuurink RC, Freitas L, et al. Gain and loss of floral scent production through changes in structural genes during pollinator-mediated speciation. Curr Biol. 2016;26(24):3303–12.

Delle-Vedove R, Schatz B, Dufay M. Understanding intraspecific variation of floral scent in light of evolutionary ecology. Ann Bot. 2017;120(1):1–20.

Pichersky E, Gershenzon J. The formation and function of plant volatiles: perfumes for pollinator attraction and defense. Curr Opin Plant Biol. 2002;5(3):237–43.

Ballerini ES, Kramer EM, Hodges SA. Comparative transcriptomics of early petal development across four diverse species of Aquilegia reveal few genes consistently associated with nectar spur development. BMC Genom. 2019;20(1):668.

Corbet SA, Willmer PG, Beament JWL, Unwin DM, Prys-Jones OE. Post-secretory determinants of sugar concentration in nectar. Plant Cell Environ. 1979;2(4):293–308.

Galliot C, Hoballah ME, Kuhlemeier C, Stuurman J. Genetics of flower size and nectar volume in Petunia pollination syndromes. Planta. 2006;225(1):203–12.

Vila-Aiub MM, Neve P, Powles SB. Fitness costs associated with evolved herbicide resistance alleles in plants. New Phytol. 2009;184(4):751–67.

Baucom RS. Evolutionary and ecological insights from herbicide-resistant weeds: what have we learned about plant adaptation, and what is left to uncover? New Phytol. 2019;223(1):68–82.

Bajwa AA, Latif S, Borger C, Iqbal N, Asaduzzaman M, Wu H, et al. The remarkable journey of a weed: biology and management of annual ryegrass ( Lolium rigidum ) in conservation cropping systems of Australia. Plants (Basel). 2021;10(8):1505.

Bitarafan Z, Andreasen C. Fecundity allocation in some european weed species competing with crops. Agronomy. 2022;12(5):1196.

Costea M, Weaver SE, Tardif FJ. The biology of Canadian weeds. 130. Amaranthus retroflexus L., A. powellii , A. powellii S. Watson, and A. hybridus L. Can J Plant Sci. 2004;84(2):631–68.

Dixon A, Comont D, Slavov GT, Neve P. Population genomics of selectively neutral genetic structure and herbicide resistance in UK populations of Alopecurus myosuroides . Pest Manag Sci. 2021;77(3):1520–9.

Kersten S, Chang J, Huber CD, Voichek Y, Lanz C, Hagmaier T, et al. Standing genetic variation fuels rapid evolution of herbicide resistance in blackgrass. Proc Natl Acad Sci U S A. 2023;120(16):e2206808120.

Qiu J, Zhou Y, Mao L, Ye C, Wang W, Zhang J, et al. Genomic variation associated with local adaptation of weedy rice during de-domestication. Nat Commun. 2017;8(1):15323.

Kreiner JM, Caballero A, Wright SI, Stinchcombe JR. Selective ancestral sorting and de novo evolution in the agricultural invasion of Amaranthus tuberculatus . Evolution. 2022;76(1):70–85.

Kreiner JM, Latorre SM, Burbano HA, Stinchcombe JR, Otto SP, Weigel D, et al. Rapid weed adaptation and range expansion in response to agriculture over the past two centuries. Science. 2022;378(6624):1079–85.

Wu D, Shen E, Jiang B, Feng Y, Tang W, Lao S, et al. Genomic insights into the evolution of Echinochloa species as weed and orphan crop. Nat Commun. 2022;13(1):689.

Yeaman S, Hodgins KA, Lotterhos KE, Suren H, Nadeau S, Degner JC, et al. Convergent local adaptation to climate in distantly related conifers. Science. 2016;353(6306):1431–3.

Haudry A, Platts AE, Vello E, Hoen DR, Leclercq M, Williamson RJ, et al. An atlas of over 90,000 conserved noncoding sequences provides insight into crucifer regulatory regions. Nat Genet. 2013;45(8):891–8.

Sackton TB, Grayson P, Cloutier A, Hu Z, Liu JS, Wheeler NE, et al. Convergent regulatory evolution and loss of flight in paleognathous birds. Science. 2019;364(6435):74–8.

Ye CY, Fan L. Orphan crops and their wild relatives in the genomic era. Mol Plant. 2021;14(1):27–39.

Clements DR, Jones VL. Ten ways that weed evolution defies human management efforts amidst a changing climate. Agronomy. 2021;11(2):284.

Article   CAS   Google Scholar  

Weinig C. Rapid evolutionary responses to selection in heterogeneous environments among agricultural and nonagricultural weeds. Int J Plant Sci. 2005;166(4):641–7.

Cousens RD, Fournier-Level A. Herbicide resistance costs: what are we actually measuring and why? Pest Manag Sci. 2018;74(7):1539–46.

Lasky JR, Josephs EB, Morris GP. Genotype–environment associations to reveal the molecular basis of environmental adaptation. Plant Cell. 2023;35(1):125–38.

Lotterhos KE. The effect of neutral recombination variation on genome scans for selection. G3-Genes Genom Genet. 2019;9(6):1851–67.

Lovell JT, MacQueen AH, Mamidi S, Bonnette J, Jenkins J, Napier JD, et al. Genomic mechanisms of climate adaptation in polyploid bioenergy switchgrass. Nature. 2021;590(7846):438–44.

Todesco M, Owens GL, Bercovich N, Légaré J-S, Soudi S, Burge DO, et al. Massive haplotypes underlie ecotypic differentiation in sunflowers. Nature. 2020;584(7822):602–7.

Revolinski SR, Maughan PJ, Coleman CE, Burke IC. Preadapted to adapt: Underpinnings of adaptive plasticity revealed by the downy brome genome. Commun Biol. 2023;6(1):326.

Kuester A, Conner JK, Culley T, Baucom RS. How weeds emerge: a taxonomic and trait-based examination using United States data. New Phytol. 2014;202(3):1055–68.

Arnaud JF, Fénart S, Cordellier M, Cuguen J. Populations of weedy crop-wild hybrid beets show contrasting variation in mating system and population genetic structure. Evol Appl. 2010;3(3):305–18.

Ellstrand NC, Schierenbeck KA. Hybridization as a stimulus for the evolution of invasiveness in plants? Proc Natl Acad Sci U S A. 2000;97(13):7043–50.

Nakabayashi K, Leubner-Metzger G. Seed dormancy and weed emergence: from simulating environmental change to understanding trait plasticity, adaptive evolution, and population fitness. J Exp Bot. 2021;72(12):4181–5.

Busi R, Yu Q, Barrett-Lennard R, Powles S. Long distance pollen-mediated flow of herbicide resistance genes in Lolium rigidum . Theor Appl Genet. 2008;117(8):1281–90.

Délye C, Clément JAJ, Pernin F, Chauvel B, Le Corre V. High gene flow promotes the genetic homogeneity of arable weed populations at the landscape level. Basic Appl Ecol. 2010;11(6):504–12.

Roumet M, Noilhan C, Latreille M, David J, Muller MH. How to escape from crop-to-weed gene flow: phenological variation and isolation-by-time within weedy sunflower populations. New Phytol. 2013;197(2):642–54.

Moghadam SH, Alebrahim MT, Mohebodini M, MacGregor DR. Genetic variation of Amaranthus retroflexus L. and Chenopodium album L. (Amaranthaceae) suggests multiple independent introductions into Iran. Front Plant Sci. 2023;13:1024555.

Muller M-H, Latreille M, Tollon C. The origin and evolution of a recent agricultural weed: population genetic diversity of weedy populations of sunflower ( Helianthus annuus L.) in Spain and France. Evol Appl. 2011;4(3):499–514.

Wesse C, Welk E, Hurka H, Neuffer B. Geographical pattern of genetic diversity in Capsella bursa-pastoris (Brassicaceae) -A global perspective. Ecol Evol. 2021;11(1):199–213.

Fraimout A, Debat V, Fellous S, Hufbauer RA, Foucaud J, Pudlo P, et al. Deciphering the routes of invasion of Drosophila suzukii by means of ABC random forest. Mol Biol Evol. 2017;34(4):980–96.

CAS   PubMed   PubMed Central   Google Scholar  

Battlay P, Wilson J, Bieker VC, Lee C, Prapas D, Petersen B, et al. Large haploblocks underlie rapid adaptation in the invasive weed Ambrosia artemisiifolia . Nat Commun. 2023;14(1):1717.

van Boheemen LA, Hodgins KA. Rapid repeatable phenotypic and genomic adaptation following multiple introductions. Mol Ecol. 2020;29(21):4102–17.

Putra A, Hodgins K, Fournier-Level A. Assessing the invasive potential of different source populations of ragweed ( Ambrosia artemisiifolia L.) through genomically-informed species distribution modelling. Authorea. 2023;17(1):e13632.

Google Scholar  

Bourguet D, Delmotte F, Franck P, Guillemaud T, Reboud X, Vacher C, et al. Heterogeneity of selection and the evolution of resistance. Trends Ecol Evol. 2013;28(2):110–8.

The International Herbicide-Resistant Weed Database. . Accessed 20 June 2023.

Powles S. Herbicide discovery through innovation and diversity. Adv Weed Sci. 2022;40(spe1):e020220074.

Murphy BP, Tranel PJ. Target-site mutations conferring herbicide resistance. Plants (Basel). 2019;8(10):382.

Gaines TA, Duke SO, Morran S, Rigon CAG, Tranel PJ, Küpper A, et al. Mechanisms of evolved herbicide resistance. J Biol Chem. 2020;295(30):10307–30.

Lonhienne T, Cheng Y, Garcia MD, Hu SH, Low YS, Schenk G, et al. Structural basis of resistance to herbicides that target acetohydroxyacid synthase. Nat Commun. 2022;13(1):3368.

Comont D, MacGregor DR, Crook L, Hull R, Nguyen L, Freckleton RP, et al. Dissecting weed adaptation: fitness and trait correlations in herbicide-resistant Alopecurus myosuroides . Pest Manag Sci. 2022;78(7):3039–50.

Neve P. Simulation modelling to understand the evolution and management of glyphosate resistance in weeds. Pest Manag Sci. 2008;64(4):392–401.

Torra J, Alcántara-de la Cruz R. Molecular mechanisms of herbicide resistance in weeds. Genes (Basel). 2022;13(11):2025.

Délye C, Gardin JAC, Boucansaud K, Chauvel B, Petit C. Non-target-site-based resistance should be the centre of attention for herbicide resistance research: Alopecurus myosuroides as an illustration. Weed Res. 2011;51(5):433–7.

Chandra S, Leon RG. Genome-wide evolutionary analysis of putative non-specific herbicide resistance genes and compilation of core promoters between monocots and dicots. Genes (Basel). 2022;13(7):1171.

Margaritopoulou T, Tani E, Chachalis D, Travlos I. Involvement of epigenetic mechanisms in herbicide resistance: the case of Conyza canadensis . Agriculture. 2018;8(1):17.

Pan L, Guo Q, Wang J, Shi L, Yang X, Zhou Y, et al. CYP81A68 confers metabolic resistance to ALS and ACCase-inhibiting herbicides and its epigenetic regulation in Echinochloa crus-galli . J Hazard Mater. 2022;428:128225.

Sen MK, Hamouzová K, Košnarová P, Roy A, Soukup J. Herbicide resistance in grass weeds: Epigenetic regulation matters too. Front Plant Sci. 2022;13:1040958.

Han H, Yu Q, Beffa R, González S, Maiwald F, Wang J, et al. Cytochrome P450 CYP81A10v7 in Lolium rigidum confers metabolic resistance to herbicides across at least five modes of action. Plant J. 2021;105(1):79–92.

Kubis GC, Marques RZ, Kitamura RS, Barroso AA, Juneau P, Gomes MP. Antioxidant enzyme and Cytochrome P450 activities are involved in horseweed ( Conyza sumatrensis ) resistance to glyphosate. Stress. 2023;3(1):47–57.

Qiao Y, Zhang N, Liu J, Yang H. Interpretation of ametryn biodegradation in rice based on joint analyses of transcriptome, metabolome and chemo-characterization. J Hazard Mater. 2023;445:130526.

Rouse CE, Roma-Burgos N, Barbosa Martins BA. Physiological assessment of non–target site restistance in multiple-resistant junglerice ( Echinochloa colona ). Weed Sci. 2019;67(6):622–32.

Abou-Khater L, Maalouf F, Jighly A, Alsamman AM, Rubiales D, Rispail N, et al. Genomic regions associated with herbicide tolerance in a worldwide faba bean ( Vicia faba L.) collection. Sci Rep. 2022;12(1):158.

Gupta S, Harkess A, Soble A, Van Etten M, Leebens-Mack J, Baucom RS. Interchromosomal linkage disequilibrium and linked fitness cost loci associated with selection for herbicide resistance. New Phytol. 2023;238(3):1263–77.

Kreiner JM, Tranel PJ, Weigel D, Stinchcombe JR, Wright SI. The genetic architecture and population genomic signatures of glyphosate resistance in Amaranthus tuberculatus . Mol Ecol. 2021;30(21):5373–89.

Parcharidou E, Dücker R, Zöllner P, Ries S, Orru R, Beffa R. Recombinant glutathione transferases from flufenacet-resistant black-grass ( Alopecurus myosuroides Huds.) form different flufenacet metabolites and differ in their interaction with pre- and post-emergence herbicides. Pest Manag Sci. 2023;79(9):3376–86.

Békés M, Langley DR, Crews CM. PROTAC targeted protein degraders: the past is prologue. Nat Rev Drug Discov. 2022;21(3):181–200.

Acuner Ozbabacan SE, Engin HB, Gursoy A, Keskin O. Transient protein-protein interactions. Protein Eng Des Sel. 2011;24(9):635–48.

Lu H, Zhou Q, He J, Jiang Z, Peng C, Tong R, et al. Recent advances in the development of protein–protein interactions modulators: mechanisms and clinical trials. Signal Transduct Target Ther. 2020;5(1):213.

Benson CW, Sheltra MR, Maughan PJ, Jellen EN, Robbins MD, Bushman BS, et al. Homoeologous evolution of the allotetraploid genome of Poa annua L. BMC Genom. 2023;24(1):350.

Robbins MD, Bushman BS, Huff DR, Benson CW, Warnke SE, Maughan CA, et al. Chromosome-scale genome assembly and annotation of allotetraploid annual bluegrass ( Poa annua L.). Genome Biol Evol. 2022;15(1):evac180.

Montgomery JS, Giacomini D, Waithaka B, Lanz C, Murphy BP, Campe R, et al. Draft genomes of Amaranthus tuberculatus , Amaranthus hybridus and Amaranthus palmeri . Genome Biol Evol. 2020;12(11):1988–93.

Jeschke MR, Tranel PJ, Rayburn AL. DNA content analysis of smooth pigweed ( Amaranthus hybridus ) and tall waterhemp ( A. tuberculatus ): implications for hybrid detection. Weed Sci. 2003;51(1):1–3.

Rayburn AL, McCloskey R, Tatum TC, Bollero GA, Jeschke MR, Tranel PJ. Genome size analysis of weedy Amaranthus species. Crop Sci. 2005;45(6):2557–62.

Laforest M, Martin SL, Bisaillon K, Soufiane B, Meloche S, Tardif FJ, et al. The ancestral karyotype of the Heliantheae Alliance, herbicide resistance, and human allergens: Insights from the genomes of common and giant ragweed. Plant Genome . 2024;e20442. .

Mulligan GA. Chromosome numbers of Canadian weeds. I Canad J Bot. 1957;35(5):779–89.

Meyer L, Causse R, Pernin F, Scalone R, Bailly G, Chauvel B, et al. New gSSR and EST-SSR markers reveal high genetic diversity in the invasive plant Ambrosia artemisiifolia L. and can be transferred to other invasive Ambrosia species. PLoS One. 2017;12(5):e0176197.

Pustahija F, Brown SC, Bogunić F, Bašić N, Muratović E, Ollier S, et al. Small genomes dominate in plants growing on serpentine soils in West Balkans, an exhaustive study of 8 habitats covering 308 taxa. Plant Soil. 2013;373(1):427–53.

Kubešová M, Moravcova L, Suda J, Jarošík V, Pyšek P. Naturalized plants have smaller genomes than their non-invading relatives: a flow cytometric analysis of the Czech alien flora. Preslia. 2010;82(1):81–96.

Thébaud C, Abbott RJ. Characterization of invasive Conyza species (Asteraceae) in Europe: quantitative trait and isozyme analysis. Am J Bot. 1995;82(3):360–8.

Garcia S, Hidalgo O, Jakovljević I, Siljak-Yakovlev S, Vigo J, Garnatje T, et al. New data on genome size in 128 Asteraceae species and subspecies, with first assessments for 40 genera, 3 tribes and 2 subfamilies. Plant Biosyst. 2013;147(4):1219–27.

Zhao X, Yi L, Ren Y, Li J, Ren W, Hou Z, et al. Chromosome-scale genome assembly of the yellow nutsedge ( Cyperus esculentus ). Genome Biol Evol. 2023;15(3):evad027.

Bennett MD, Leitch IJ, Hanson L. DNA amounts in two samples of angiosperm weeds. Ann Bot. 1998;82:121–34.

Schulz-Schaeffer J, Gerhardt S. Cytotaxonomic analysis of the Euphorbia spp. (leafy spurge) complex. II: Comparative study of the chromosome morphology. Biol Zentralbl. 1989;108(1):69–76.

Schaeffer JR, Gerhardt S. The impact of introgressive hybridization on the weediness of leafy spurge. Leafy Spurge Symposium. 1989;1989:97–105.

Bai C, Alverson WS, Follansbee A, Waller DM. New reports of nuclear DNA content for 407 vascular plant taxa from the United States. Ann Bot. 2012;110(8):1623–9.

Aarestrup JR, Karam D, Fernandes GW. Chromosome number and cytogenetics of Euphorbia heterophylla L. Genet Mol Res. 2008;7(1):217–22.

Wang L, Sun X, Peng Y, Chen K, Wu S, Guo Y, et al. Genomic insights into the origin, adaptive evolution, and herbicide resistance of Leptochloa chinensis , a devastating tetraploid weedy grass in rice fields. Mol Plant. 2022;15(6):1045–58.

Paril J, Pandey G, Barnett EM, Rane RV, Court L, Walsh T, et al. Rounding up the annual ryegrass genome: high-quality reference genome of Lolium rigidum . Front Genet. 2022;13:1012694.

Weiss-Schneeweiss H, Greilhuber J, Schneeweiss GM. Genome size evolution in holoparasitic Orobanche (Orobanchaceae) and related genera. Am J Bot. 2006;93(1):148–56.

Towers G, Mitchell J, Rodriguez E, Bennett F, Subba Rao P. Biology & chemistry of Parthenium hysterophorus L., a problem weed in India. Biol Rev. 1977;48:65–74.

CAS   Google Scholar  

Moghe GD, Hufnagel DE, Tang H, Xiao Y, Dworkin I, Town CD, et al. Consequences of whole-genome triplication as revealed by comparative genomic analyses of the wild radish ( Raphanus raphanistrum ) and three other Brassicaceae species. Plant Cell. 2014;26(5):1925–37.

Zhang X, Liu T, Wang J, Wang P, Qiu Y, Zhao W, et al. Pan-genome of Raphanus highlights genetic variation and introgression among domesticated, wild, and weedy radishes. Mol Plant. 2021;14(12):2032–55.

Chytrý M, Danihelka J, Kaplan Z, Wild J, Holubová D, Novotný P, et al. Pladias database of the Czech flora and vegetation. Preslia. 2021;93(1):1–87.

Patterson EL, Pettinga DJ, Ravet K, Neve P, Gaines TA. Glyphosate resistance and EPSPS gene duplication: Convergent evolution in multiple plant species. J Hered. 2018;109(2):117–25.

Jugulam M, Niehues K, Godar AS, Koo DH, Danilova T, Friebe B, et al. Tandem amplification of a chromosomal segment harboring 5-enolpyruvylshikimate-3-phosphate synthase locus confers glyphosate resistance in Kochia scoparia . Plant Physiol. 2014;166(3):1200–7.

Patterson EL, Saski CA, Sloan DB, Tranel PJ, Westra P, Gaines TA. The draft genome of Kochia scoparia and the mechanism of glyphosate resistance via transposon-mediated EPSPS tandem gene duplication. Genome Biol Evol. 2019;11(10):2927–40.

Zhang C, Johnson N, Hall N, Tian X, Yu Q, Patterson E. Subtelomeric 5-enolpyruvylshikimate-3-phosphate synthase ( EPSPS ) copy number variation confers glyphosate resistance in Eleusine indica . Nat Commun. 2023;14:4865.

Koo D-H, Molin WT, Saski CA, Jiang J, Putta K, Jugulam M, et al. Extrachromosomal circular DNA-based amplification and transmission of herbicide resistance in crop weed Amaranthus palmeri . Proc Natl Acad Sci U S A. 2018;115(13):3332–7.

Molin WT, Yaguchi A, Blenner M, Saski CA. The eccDNA Replicon: A heritable, extranuclear vehicle that enables gene amplification and glyphosate resistance in Amaranthus palmeri . Plant Cell. 2020;32(7):2132–40.

Jugulam M. Can non-Mendelian inheritance of extrachromosomal circular DNA-mediated EPSPS gene amplification provide an opportunity to reverse resistance to glyphosate? Weed Res. 2021;61(2):100–5.

Kreiner JM, Giacomini DA, Bemm F, Waithaka B, Regalado J, Lanz C, et al. Multiple modes of convergent adaptation in the spread of glyphosate-resistant Amaranthus tuberculatus . Proc Natl Acad Sci U S A. 2019;116(42):21076–84.

Cai L, Comont D, MacGregor D, Lowe C, Beffa R, Neve P, et al. The blackgrass genome reveals patterns of non-parallel evolution of polygenic herbicide resistance. New Phytol. 2023;237(5):1891–907.

Chen K, Yang H, Peng Y, Liu D, Zhang J, Zhao Z, et al. Genomic analyses provide insights into the polyploidization-driven herbicide adaptation in Leptochloa weeds. Plant Biotechnol J. 2023;21(8):1642–58.

Ohadi S, Hodnett G, Rooney W, Bagavathiannan M. Gene flow and its consequences in Sorghum spp. Crit Rev Plant Sci. 2017;36(5–6):367–85.

Renzi JP, Coyne CJ, Berger J, von Wettberg E, Nelson M, Ureta S, et al. How could the use of crop wild relatives in breeding increase the adaptation of crops to marginal environments? Front Plant Sci. 2022;13:886162.

Ward SM, Cousens RD, Bagavathiannan MV, Barney JN, Beckie HJ, Busi R, et al. Agricultural weed research: a critique and two proposals. Weed Sci. 2014;62(4):672–8.

Evans JA, Tranel PJ, Hager AG, Schutte B, Wu C, Chatham LA, et al. Managing the evolution of herbicide resistance. Pest Manag Sci. 2016;72(1):74–80.

International Weed Genomics Consortium Website. . Accessed 20 June 2023.

WeedPedia Database. . Accessed 20 June 2023.

Hall N, Chen J, Matzrafi M, Saski CA, Westra P, Gaines TA, et al. FHY3/FAR1 transposable elements generate adaptive genetic variation in the Bassia scoparia genome. bioRxiv . 2023; DOI: .

Jarvis DE, Sproul JS, Navarro-Domínguez B, Krak K, Jaggi K, Huang Y-F, et al. Chromosome-scale genome assembly of the hexaploid Taiwanese goosefoot “Djulis” ( Chenopodium formosanum ). Genome Biol Evol. 2022;14(8):evac120.

Ferreira LAI, de Oliveira RS, Jr., Constantin J, Brunharo C. Evolution of ACCase-inhibitor resistance in Chloris virgata is conferred by a Trp2027Cys mutation in the herbicide target site. Pest Manag Sci. 2023;79(12):5220–9.

Laforest M, Martin SL, Bisaillon K, Soufiane B, Meloche S, Page E. A chromosome-scale draft sequence of the Canada fleabane genome. Pest Manag Sci. 2020;76(6):2158–69.

Guo L, Qiu J, Ye C, Jin G, Mao L, Zhang H, et al. Echinochloa crus-galli genome analysis provides insight into its adaptation and invasiveness as a weed. Nat Commun. 2017;8(1):1031.

Sato MP, Iwakami S, Fukunishi K, Sugiura K, Yasuda K, Isobe S, et al. Telomere-to-telomere genome assembly of an allotetraploid pernicious weed, Echinochloa phyllopogon . DNA Res. 2023;30(5):dsad023.

Stein JC, Yu Y, Copetti D, Zwickl DJ, Zhang L, Zhang C, et al. Genomes of 13 domesticated and wild rice relatives highlight genetic conservation, turnover and innovation across the genus Oryza . Nat Genet. 2018;50(2):285–96.

Wu D, Xie L, Sun Y, Huang Y, Jia L, Dong C, et al. A syntelog-based pan-genome provides insights into rice domestication and de-domestication. Genome Biol. 2023;24(1):179.

Wang Z, Huang S, Yang Z, Lai J, Gao X, Shi J. A high-quality, phased genome assembly of broomcorn millet reveals the features of its subgenome evolution and 3D chromatin organization. Plant Commun. 2023;4(3):100557.

Mao Q, Huff DR. The evolutionary origin of Poa annua L. Crop Sci. 2012;52(4):1910–22.

Benson CW, Sheltra MR, Maughan JP, Jellen EN, Robbins MD, Bushman BS, et al. Homoeologous evolution of the allotetraploid genome of Poa annua L. Res Sq. 2023. .

Brunharo C, Benson CW, Huff DR, Lasky JR. Chromosome-scale genome assembly of Poa trivialis and population genomics reveal widespread gene flow in a cool-season grass seed production system. Plant Direct. 2024;8(3):e575.

Mo C, Wu Z, Shang X, Shi P, Wei M, Wang H, et al. Chromosome-level and graphic genomes provide insights into metabolism of bioactive metabolites and cold-adaption of Pueraria lobata var. montana . DNA Research. 2022;29(5):dsac030.

Thielen PM, Pendleton AL, Player RA, Bowden KV, Lawton TJ, Wisecaver JH. Reference genome for the highly transformable Setaria viridis ME034V. G3 (Bethesda, Md). 2020;10(10):3467–78.

Yoshida S, Kim S, Wafula EK, Tanskanen J, Kim Y-M, Honaas L, et al. Genome sequence of Striga asiatica provides insight into the evolution of plant parasitism. Curr Biol. 2019;29(18):3041–52.

Qiu S, Bradley JM, Zhang P, Chaudhuri R, Blaxter M, Butlin RK, et al. Genome-enabled discovery of candidate virulence loci in Striga hermonthica , a devastating parasite of African cereal crops. New Phytol. 2022;236(2):622–38.

Nunn A, Rodríguez-Arévalo I, Tandukar Z, Frels K, Contreras-Garrido A, Carbonell-Bejerano P, et al. Chromosome-level Thlaspi arvense genome provides new tools for translational research and for a newly domesticated cash cover crop of the cooler climates. Plant Biotechnol J. 2022;20(5):944–63.

USDA-ARS Germplasm Resources Information Network (GRIN). . Accessed 20 June 2023.

Buck M, Hamilton C. The Nagoya Protocol on access to genetic resources and the fair and equitable sharing of benefits arising from their utilization to the convention on biological diversity. RECIEL. 2011;20(1):47–61.

Chauhan BS, Matloob A, Mahajan G, Aslam F, Florentine SK, Jha P. Emerging challenges and opportunities for education and research in weed science. Front Plant Sci. 2017;8:1537.

Shah S, Lonhienne T, Murray CE, Chen Y, Dougan KE, Low YS, et al. Genome-guided analysis of seven weed species reveals conserved sequence and structural features of key gene targets for herbicide development. Front Plant Sci. 2022;13:909073.

International Weed Genomics Consortium Training Resources. . Accessed 20 June 2023.

Blackford S. Harnessing the power of communities: career networking strategies for bioscience PhD students and postdoctoral researchers. FEMS Microbiol Lett. 2018;365(8):fny033.

Pender M, Marcotte DE, Sto Domingo MR, Maton KI. The STEM pipeline: The role of summer research experience in minority students’ Ph.D. aspirations. Educ Policy Anal Arch. 2010;18(30):1–36.

PubMed   PubMed Central   Google Scholar  

Burke A, Okrent A, Hale K. The state of U.S. science and engineering 2022. Foundation NS. . 2022.

Wu J-Y, Liao C-H, Cheng T, Nian M-W. Using data analytics to investigate attendees’ behaviors and psychological states in a virtual academic conference. Educ Technol Soc. 2021;24(1):75–91.

Download references

Peer review information

Wenjing She was the primary editor of this article and managed its editorial process and peer review in collaboration with the rest of the editorial team.

The International Weed Genomics Consortium is supported by BASF SE, Bayer AG, Syngenta Ltd, Corteva Agriscience, CropLife International (Global Herbicide Resistance Action Committee), the Foundation for Food and Agriculture Research (Award DSnew-0000000024), and two conference grants from USDA-NIFA (Award numbers 2021–67013-33570 and 2023-67013-38785).

Author information

Authors and affiliations.

Department of Agricultural Biology, Colorado State University, 1177 Campus Delivery, Fort Collins, CO, 80523, USA

Jacob Montgomery, Sarah Morran & Todd A. Gaines

Protecting Crops and the Environment, Rothamsted Research, Harpenden, Hertfordshire, UK

Dana R. MacGregor

Department of Crop, Soil, and Environmental Sciences, Auburn University, Auburn, AL, USA

J. Scott McElroy

Department of Plant and Environmental Sciences, University of Copenhagen, Taastrup, Denmark

Paul Neve & Célia Neto

IFEVA-Conicet-Department of Ecology, University of Buenos Aires, Buenos Aires, Argentina

Martin M. Vila-Aiub & Maria Victoria Sandoval

Department of Ecology, Faculty of Agronomy, University of Buenos Aires, Buenos Aires, Argentina

Analia I. Menéndez

Department of Botany, The University of British Columbia, Vancouver, BC, Canada

Julia M. Kreiner

Institute of Crop Sciences, Zhejiang University, Hangzhou, China

Longjiang Fan

Department of Biology, University of Massachusetts Amherst, Amherst, MA, USA

Ana L. Caicedo

Department of Plant and Wildlife Sciences, Brigham Young University, Provo, UT, USA

Peter J. Maughan

Bayer AG, Weed Control Research, Frankfurt, Germany

Bianca Assis Barbosa Martins, Jagoda Mika, Alberto Collavo & Bodo Peters

Department of Crop Sciences, Federal University of Rio Grande Do Sul, Porto Alegre, Rio Grande Do Sul, Brazil

Aldo Merotto Jr.

Department of Soil and Crop Sciences, Texas A&M University, College Station, TX, USA

Nithya K. Subramanian & Muthukumar V. Bagavathiannan

Department of Plant, Soil and Microbial Sciences, Michigan State University, East Lansing, MI, USA

Luan Cutti & Eric L. Patterson

Department of Agronomy, Kansas State University, Manhattan, KS, USA

Md. Mazharul Islam & Mithila Jugulam

Department of Plant Pathology, Kansas State University, Manhattan, KS, USA

Bikram S. Gill

Crop Protection Discovery and Development, Corteva Agriscience, Indianapolis, IN, USA

Robert Cicchillo, Roger Gast & Neeta Soni

Genome Center of Excellence, Corteva Agriscience, Johnston, IA, USA

Terry R. Wright, Gina Zastrow-Hayes, Gregory May, Kevin Fengler & Victor Llaca

School of Agriculture, Food and Wine, University of Adelaide, Glen Osmond, South Australia, Australia

Jenna M. Malone

Jealott’s Hill International Research Centre, Syngenta Ltd, Bracknell, Berkshire, UK

Deepmala Sehgal, Shiv Shankhar Kaundun & Richard P. Dale

Department of Plant and Soil Sciences, University of Pretoria, Pretoria, South Africa

Barend Juan Vorster

BASF SE, Ludwigshafen Am Rhein, Germany

Jens Lerchl

Department of Crop Sciences, University of Illinois, Urbana, IL, USA

Patrick J. Tranel

Senior Scientist Consultant, Herbicide Resistance Action Committee / CropLife International, Liederbach, Germany

Roland Beffa

School of BioSciences, University of Melbourne, Parkville, VIC, Australia

Alexandre Fournier-Level

You can also search for this author in PubMed   Google Scholar


JMo and TG conceived and outlined the article. TG, DM, EP, RB, JSM, PJT, MJ wrote grants to obtain funding. MMI, BSG, and MJ performed mitotic chromosome visualization. VL performed sequencing. VL and KF assembled the genomes. LC and ELP annotated the genomes. JMo, SM, DRM, JSM, PN, CN, MV, MVS, AIM, JMK, LF, ALC, PJM, BABM, JMi, AC, MVB, LC, AFL, and ELP wrote the first draft of the article. All authors edited the article and improved the final version.

Corresponding author

Correspondence to Todd A. Gaines .

Ethics declarations

Ethics approval and consent to participate.

Ethical approval is not applicable for this article.

Competing interests

Some authors work for commercial agricultural companies (BASF, Bayer, Corteva Agriscience, or Syngenta) that develop and sell weed control products.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information


Additional file 1. List of completed and in-progress genome assemblies of weed species pollinated by insects (Table S1).


Additional file 2. Methods and results for visualizing and counting the metaphase chromosomes of hexaploid Avena fatua (Fig S1); diploid Lolium rigidum  (Fig S2); tetraploid Phalaris minor (Fig S3); and tetraploid Salsola tragus (Fig S4).

Additional file 3. Review history.

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit . The Creative Commons Public Domain Dedication waiver ( ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Montgomery, J., Morran, S., MacGregor, D.R. et al. Current status of community resources and priorities for weed genomics research. Genome Biol 25 , 139 (2024).

Download citation

Received : 11 July 2023

Accepted : 13 May 2024

Published : 27 May 2024


Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

  • Weed science
  • Reference genomes
  • Rapid adaptation
  • Herbicide resistance
  • Public resources

Genome Biology

ISSN: 1474-760X

research methods for communication


  1. Modern Methods of Communication

    research methods for communication

  2. Quantitative Research Methods for Communication: A Hands-On Approach

    research methods for communication

  3. Communication Research Methods

    research methods for communication

  4. PPT

    research methods for communication

  5. Communication Research Methods 4th Edition

    research methods for communication

  6. PPT

    research methods for communication


  1. Communication Research Methods

  2. Research Methodology 101: Simple Explainer With Examples (+ FREE Template)

  3. Communicating Research Findings [Video 7]

  4. Three Paradigms of Communication Research

  5. Communication Research Methods

  6. Introduction to Communication research|| Unit 10|| NTA UGC NET MASS COMMUNICATION|| By Priyanka rana


  1. Guide to Communication Research Methodologies: Quantitative

    Imagine two researchers who want to understand student perceptions of the basic communication course at a university. The first researcher, a quantitative researcher, might measure absences to understand student perception. The second researcher, a qualitative researcher, might interview students to find out what they like and dislike about a ...

  2. Research Methods

    Mass Communications Research Methods. Originally published in 1988. Step-by-step, this book leads students from problem identification, through the mazes of surveys, experimentation, historical/qualitative studies, statistical analysis, and computer data processing to the final submission and publication in scientific or popular publications.

  3. Media and Communication Research Methods

    The Fifth Edition of Media and Communication Research Methods includes a new chapter on discourse analysis; expanded discussion of social media, including discussion of the ethics of Facebook experiments; and expanded coverage of the research process with new discussion of search strategies and best practices for analyzing research articles.

  4. Understanding Communication Research Methods

    The book features all the main research traditions within communication including online methods and provides level-appropriate applications of the methods through theoretical and practical examples and exercises, including sample student papers that demonstrate research methods in action. This third edition also includes additional chapters on ...

  5. Qualitative Communication Research Methods

    Qualitative Communication Research Methods, Fourth Edition introduces readers to qualitative research in speech and mass communication. Award-winning scholars and authors Thomas R. Lindlof and Bryan C. Taylor guide readers through every step of the qualitative process, from developing research topics and questions through writing a final report

  6. Three Types of Communication Research Methods: Quantitative ...

    All research begins with some set of assumptions which themselves are untested but believed.Positivistic research, which comprises the mass of modern communication and development research, proceeds from the presupposition that all knowledge is based on an observable reality and social phenomena can be studied on the basis of methodologies and techniques adopted from the natural sciences.

  7. Taking communication science and research methodology seriously

    In addition, new challenges and issues have emerged for CMM since the introduction of computational methods to the social sciences (e.g., Lazer et al., Citation 2009) and the communication discipline in particular (e.g., van Atteveldt & Peng, Citation 2018).Computational methods are new and still fast developing, and interdisciplinary. There also might be a lack of a clear benchmark for what ...

  8. Quantitative research methods for communication: a hands-on approach

    Quantitative research methods for communication: a hands-on approach by Jason S. Wrench, Candice Thomas-Maddox, Virginia Peck Richmond and James C. McCroskey, New York, NY, Oxford University Press, 2018, (4th edition), 672 pp., $94.95 (paperback), ISBN 9780190861063

  9. Research Methods in Intercultural Communication: A Practical Guide

    Fred Dervin is Professor of Multicultural Education at the University of Helsinki (Finland). He also holds several associate professorships around the world. specializes in language and intercultural education, the sociology of multicultural-ism, and linguistics for intercultural communication and education.

  10. Quantitative Research Methods for Communication: A Hands-On Approach

    Conducting research can be one of the most fascinating--and intimidating--tasks for students and scholars. A relevant and accessible guide to quantitative research, Quantitative Research Methods for Communication: A Hands-On Approach offers an innovative and insightful look at this complex subject.

  11. Methods for Intercultural Communication Research

    Summary. Research on intercultural communication is conducted using primarily three different methodological approaches: social scientific, interpretive, and critical. Each of these approaches reflects different philosophical assumptions about the world and how we come to know it. Social scientific methods often involve quantitative data ...

  12. Quantitative Research Methods for Communication: A Hands-On Approach 4e

    Description. Quantitative Research Methods for Communication: A Hands-On Approach is a comprehensive and engaging guide to quantitative research. Drawing on their extensive research and teaching experience in the field of communication, authors Jason S. Wrench, Candice Thomas-Maddox, Virginia Peck Richmond, and James C. McCroskey encourage ...

  13. PDF Introduction to Communication Research Methods COM 365 Course Syllabus

    COM 365 Course Syllabus ~ Fall 2010. "The best research is driven by theory, validates a theory, further explains a theory, challenges an existing theory, or aids in the creation of theory. Theoretically driven research is built on the results of previous researchers, and it provides a foundation for subsequent researchers.

  14. (PDF) Introduction to Communication Research

    conducting communication research is an enormous. field of study that requires very skilled and knowl-. edgeable researchers who are aware of its many. areas, approaches, designs, methods, tools ...

  15. The Sage Handbook of Qualitative Research in Organizational Communication

    The Sage Handbook of Qualitative Research in Organizational Communication is a state-of-the-art resource for scholars, students, and practitioners seeking to deepen their understanding and expertise in this dynamic field. Written by a global team of established and emerging experts, this Handbook provides a comprehensive exploration of the field's foundational traditions of epistemology and ...

  16. Quantitative Research Methods in Communication

    ABSTRACT. This textbook is an advanced introduction to quantitative methods for students in communication and allied social science disciplines that focuses on why and how to conduct research that contributes to social justice. Today's researchers are inspired by the potential for scholarship to make a difference for society, to push toward ...

  17. PDF Communication Research Methods 1

    Purpose of Course: This course is designed to introduce you to the scientific, interpretive, rhetorical, and critical research methods we use to study communication problems and processes. We will focus on the critical analysis, evaluation, and use of quantitative and qualitative communication research methods.

  18. Home

    Quantitative Research Methods for Communication: A Hands-On Approach by Jason Wrench, Candice Thomas-Maddox, Virginia Richmond, and James McCroskey is currently in the 3rd edition and the 4th edition is arriving in 2019. The textbook's website contains a number of helpful video tutorials and links that help new researchers explore the world of ...

  19. Contemporary approaches to assessing mediation in communication research

    The authors spend a majority of this chapter discussing strategies by which mediation hypotheses may be formally tested statistically. They also discuss useful extensions of these strategies, such as how to address hypotheses that involve both mediation and moderation effects. They suggest ways to proceed when theory implies that multiple mediators may intervene between two variables, and ...

  20. (PDF) Research Methods in Communication

    by W. Lawrence Neuman and Karen Robson. Research Methods, Design, and Analysis, Eleventh Edition. by Larry B. Christensen, R. Burke Johnson, and Lisa A. T urner. Research Methods in. Communication ...

  21. What is Communication Research?

    The book "Mass Communication Research Methods", released in 1998, helped cement this as standard, defining the experimental methods of the day [6]. These research methods - focus groups, observations, and surveys - have now long been central to the field, yet the next step in empirical quantification is already emerging. Continuing with ...

  22. Quantitative Research Methods for Communication: A Hands-On Approach

    Quantitative Research Methods for Communication: A Hands-On Approach is a relevant and accessible guide to quantitative research. In clear, straightforward language, the authors encourage students to take an active, hands-on role in the learning process, giving them the tools they need to locate, conduct, collect and present their research.

  23. COM5104

    This course is designed to provide training on research methods in communication and new media. It aims to: oral and written presentation of research results. By the end of the course, students are expected to be able to read and evaluate professional and academic research reports; design and implement research projects; perform statistical ...

  24. Review of Muslim Patient Needs and Its Implications on Healthcare

    Leininger's theory of cultural care emphasizes the importance of providing culturally sensitive care through acknowledging and addressing patients' beliefs. 1 One approach to delivering this multi-factorial care involves cultural humility—a process of self-awareness and ongoing learning about diverse cultures. Cultural humility serves as a useful framework for healthcare providers when ...

  25. Business Communications in Project Management of International

    Purpose. The aim of this study is to investigate the impact of business communications on project management in international companies. The study aims to identify the main dependencies and provide recommendations for their effective application in company activities. Methodology of research. The study employed various scientific methods, including the analytical method for literature review ...

  26. HeteroTCR: A heterogeneous graph neural network-based method for

    HeteroTCR extracts information on within-type (TCR-TCR or peptide-peptide) similarity and between-type (peptide-TCR) interaction from peptides and TCR input. It is robust across diverse datasets ...

  27. A realist review of health passports for Autistic adults.

    Background: Autism is a normal part of cognitive diversity, resulting in communication and sensory processing differences, which can become disabling in a neurotypical world. Autistic people have an increased likelihood of physical and mental co-occurring conditions and die earlier than neurotypical peers. Inaccessible healthcare may contribute to this. Autism Health Passports (AHPs) are paper ...

  28. NAD Supplement: Benefits, Side Effects, and Efficacy

    Learn more about potential research-supported NAD benefits, including increased energy and cellular repair, to discover whether NAD is right for you. ... making things happen and maximizing communication between cells throughout the body. Without enough NAD, simple processes in the body may take longer, partly because of poor communication ...

  29. Current status of community resources and priorities for weed genomics

    Weeds are attractive models for basic and applied research due to their impacts on agricultural systems and capacity to swiftly adapt in response to anthropogenic selection pressures. Currently, a lack of genomic information precludes research to elucidate the genetic basis of rapid adaptation for important traits like herbicide resistance and stress tolerance and the effect of evolutionary ...

  30. Adaptive attack recognition method based on probability model for

    Traditional attack detection methods, relying on static anomaly thresholds from data distributions, falter in complex driving scenarios. Recent studies have sought more nuanced detection techniques, including deep learning and inter-vehicle communication, but these approaches face limitations related to data security and environmental ...