Orit Shaer, Oded Nov
Recent advances in genetic testing and Internet technologies have led to a dramatic increase in the access non-experts have to their own personal genomic information. As a result, individuals are confronted with an unprecedented amount of sensitive information about themselves, which influences their decisions, emotional state, and well-being [1,2]. The use of Web-based interactive technologies to deliver such information raises questions about how people make sense of, engage with, and rely on their personal genomic data. Such questions are not only of paramount importance for society and policy makers but are also a pressing issue for human-computer interaction (HCI) research.
Future progress in genetic research and technologies is likely to further increase the availability of interactive personal genomic information to non-expert users. This trend raises technological, ethical, and regulatory concerns. Only recently, the U.S. Food and Drug Administration (FDA) ordered that 23andMe, a direct-to-consumer genetic testing company, to stop providing risk assessment reports, claiming that “serious concerns are raised if test results are not adequately understood” [3,4].
As genomic information available to users is likely to become ever more detailed and complex, HCI tools and practices can help make it more accessible and understandable. In our view, understanding, informing, and empowering non-experts’ interaction with personal genomics is one of the key challenges that lie ahead for the HCI community.
Here, we provide a brief overview of personal genomics and explore the roles HCI can play in helping personal genomic information users to understand, engage with, and share their information. This is also a call to action for those of us interested in the intersection of personal genomics and HCI, and, more broadly, the interaction of non-experts with scientific data.
Online Interaction with Personal Genomics
In April 2003, the Human Genome Project (HGP) published the full reference sequence of the human genome. This collaborative research program, whose goal was the complete mapping and understanding of all the human genes, lasted 13 years and cost $2.7 billion. Since then, the cost of sequencing a single human genome has dropped significantly.
The decline in the costs of DNA sequencing offers the promise of personalized medicine, with genomic information integrated into medical care to provide individualized risk assessment, tailored lifestyle-change recommendations, and medications to reduce risk. In addition, it led to widespread access to personal genomics data. Several companies now offer personal genomics services directly to consumers.
Direct-to-consumer genetic testing (DTCGT) is a relatively new and developing online service that enables individuals to acquire genetic information without the mandatory involvement of a healthcare provider by sending a saliva sample to a DTCGT company—at the cost of a few hundred dollars. Several popular DTCGT services offer interactive online reports of non-health-related information including traits and ancestry information (e.g., AncestryDNA and FamilyTreeDNA). 23andMe  also provided risk-assessment results for about 250 medical conditions. However, the reporting of health-related information directly to consumers has been stopped while it is undergoing review by the FDA, which is examining whether tests results are accurate and are adequately communicated to and understood by consumers .
As genomic information available to users is likely to become ever more detailed and complex, HCI tools and practices can help make it more accessible and understandable.
At the other end of the spectrum from the commercial offering of genetic services are large-scale government and nonprofit efforts. For example, the U.K. government recently announced its plan to sequence and return whole personal genomes to 100,000 British citizens by 2017. In the U.S., the Veterans Administration is pursuing an effort to enroll one million veterans in a research study that incorporates genetic profiling. A prominent nonprofit example is PersonalGenomes.org  and their Global Network of Personal Genome Projects, a program spanning four countries that seeks to improve the scientific understanding of the genetic and environmental contributions to human traits through the creation of a genetic public database of 100,000 volunteers. Participants must be willing to share their genomic sequences as well as health and medical data with the scientific community and the general public. The longest running PGP site is based out of George Church’s Lab at Harvard Medical School. The Harvard PGP was established in 2005, when the cost of sequencing a single human genome was in excess of $1 million. It began with a pilot study of 10 fully identified individuals, known as the PGP-10, and slowly scaled up. Today, more than 3,000 U.S. citizens are enrolled through a process of open consent in the project and share their genomic information.
In summary, given recent advances in the field of personal genomics and rapidly declining sequencing costs, it seems unavoidable that people will continue to seek learning about their own genetic makeup and its health implications. Our goal is to contribute to developing interactive systems in order to help people engage with and learn from their personal genomic information.
Implications for HCI
The highly personal and complex nature of personal genomic information raises important HCI questions about how people make sense of and engage with it, and how comfortable they feel about sharing it in order to advance scientific and biomedical research. More specifically, the dynamic nature of personal genomic information, which constantly gets updated thanks to new research findings, raises the following important HCI questions:
- What are the functional requirements for supporting meaningful engagement of non-experts with personal genomic information?
- How can we design effective interaction for personal genomic information?
- How can we evaluate the effectiveness of techniques for interaction with personal genomic information?
- Can user interface design interventions motivate users’ willingness to adapt healthier behaviors?
- Can user interface design interventions impact users’ willingness to share their personal genomic data?
In the following discussion, we begin to address these questions by considering user perspectives on personal genomics. We then highlight areas for future HCI work.
User Perspectives on Personal Genomics
Little empirical data exists about the attitudes and motivations of personal genomics users . Few studies have recruited participants who had actually used genetic testing. In these studies, curiosity was mentioned as participants’ primary motivation for undergoing genetic testing [1,2]. Participants also stated that they would use information gained from the test to take personal responsibility for their future health . Other themes include fascination with genealogy, contribution to research, and recreation [1,2]. Studies also identified several concerns among personal genomics users, including privacy, the nature of the results, and their future impact [1,2,7]. Only a small number of users had their entire genomes sequenced, and to our knowledge no studies have investigated the perspectives of such users.
While the studies we described shed some light on the motivations and concerns of early personal genomics users, they do not provide insight into the information practices and needs of such users.
We began to study users’ information practices and needs by conducting a qualitative study with Personal Genome Project (PGP) participants, who are early adopters of personal genomics technologies.
Sixty-three participants (29 women) filled out an online questionnaire that consists of 10 open questions (see Table 1), in addition to demographic questions. Participants were between the ages of 21 and 71, with an average age of 47. All participants had prior access to their personal genomic data. Some participants had personal genomics information from more than one service provider. Eleven participants had their entire genomes sequenced.
Fifty-two participants held undergraduate or graduate degrees; 20 participants held doctoral degrees; and 19 participants worked in life sciences-related fields. These findings indicate that our study participants fit the description of early adopters: They tend to have advanced education and possess favorable attitudes toward science.
We analyzed participants’ responses using content-analysis methods. We developed first-level codes on the basis of existing literature and a preliminary review of the data by two independent coders, and then collapsed them into advanced categories based on frequency. Categories were analyzed for the identification of themes.
Next, we discuss findings related to two aspects of personal genomics where we see opportunities for effective HCI interventions: engaging and learning from personal genomic data and sharing personal genomic information.
Engaging and Learning from Personal Genomics
To get insight into the information needs of personal genomics users, we asked participants about the websites and computational tools they used for engaging with their personal genomic information and how they used these tools to learn from their data.
We found that online tools for personal genomics used by our participants could be classified into three general categories:
- interpretive tools, which apply computational algorithms for finding biological meaning and for annotating genomic data;
- testing services, which enable users to order DNA test kits, provide raw genetic data for download, and include some interpretative tools; and
- databases, which document SNPs (variations) that have been reported as medically or genealogically significant, and cite related scientific publications.
Although our study participants were motivated by a diverse set of goals, ranging from understanding traits to identifying health risks to learning about their ancestry, they used interactive tools to perform five common information tasks (Figure 1):
- Review an annotated report—“get an overview of my genetic makeup”
- Search the literature—“learn about specific SNPs”
- Share information—“share with genetic cousins”
- Compare genomes—“understand connections within families”
- Curate information—“compile a list of shared chromosome segments.”
We also asked participants what could help them learn more from their personal genomic data information. The following needs have emerged from user responses (also summarized in Figure 2):
- Integrated resources. Multiple users highlighted a need for the seamless integration of data resources, including annotated genomes, publications, various public databases, and health-related data. In the words of one user, “Integrated databases of published research that allow the end user, through a seamless interface, to connect personal data with any possibly relevant literature and public data.”
- Visualization and information presentation. Users also commented on the need to visualize their genomes. One user stated, “I’d be interested in seeing a graphic illustration of my chromosome sets.” Other users commented on the difficulty of interpreting tabular and dense reports.
- Data triangulation. Several users asked for the ability to triangulate data from several individuals in order to understand connections within families.
- Content. Multiple users asked for the content of personal genomic reports to be more easy to use by laypeople. Users also asked for educational materials to be integrated within the reports: “Every time I try to understand something, I have to educate myself via Google, instead of the interface that gives me my genetic data educating me. The research it takes holds me back from using my info more.” Finally, users asked for highlighting actionable information: “Features that show more clearly what reasonable actionable options there might be for dealing with or preventing various illnesses.”
- Sharing. Users also highlighted the need for tools that facilitate and encourage information sharing. For example, one user suggested that “easy-to-use at-home programs will be needed to compare one’s data with those of friends.”
These findings identify common information tasks and highlight information needs to be addressed by interactive tools for personal genomics.
From an HCI perspective, these findings can serve as a basis toward the design of new interaction techniques and tools for personal genomics. However, many open questions remain regarding non-expert engagement with personal genomic information. For example, what makes the genomic information difficult to understand? How do users find and decide which bioinformatics tools to use? How do users document and organize their findings? How do they validate their findings? Future HCI research can gain further insights on user engagement with personal genomics and play an important role in the design of new interactive tools that help non-experts to engage and learn from their personal genomic information.
Sharing Personal Genomics Data
From a societal perspective, another important aspect of users’ engagement with their personal genomic information is information sharing.
All of our study participants shared their genomic data with the PGP research. Such sharing increases the value of genomic research, as it enables researchers to study the links between genetics and health in finer detail. From an HCI perspective, it is important to understand the factors that are related to information sharing. Understanding user concerns about genomic information, and about how these can be alleviated, can contribute by (1) creating new sharing mechanisms that alleviate user concerns, for example by allowing users to control what kind of information will be shared and who it can be shared with (i.e., which research projects); (2) designing more relevant and appropriate consent forms; (3) encouraging users to share their information with researchers, given that their privacy needs are met.
The Road Ahead
Given the different ways users engage with their personal genomic information, reports could be more personalized and interactive and cater to different users’ needs. Personal genomics data already contains information that can be used to inform the presentation to the user; for example, most genetic-testing companies collect other demographic information such as age and education level. Future work may explore the utility of the automatic adaptation of interactive genetic reports based on the personal attributes of the user. Down the line, HCI research may also consider factors affecting users’ willingness to share information. For example, what kind of information do users share and with whom? What concerns do users have regarding the sharing of personal genomic information or meta-information? When people understand the information better, are they willing to share more or less?
Further work is needed to better understand the ways in which HCI techniques can contribute to engaging with, learning from, and sharing personal genomic information more effectively.
Overall, bringing HCI knowledge to the growing field of personal genomics can help make two important contributions: improve genetic literacy among non-experts and improve the facilitation of sharing personal genomic information. Both are important for a more informed and healthy society.
1. Gollust, S.E., Gordon, E.S., Zayac, C., Griffin, G., Christman, M.F., Pyeritz, R.E. et al. Motivations and perceptions of early adopters of personalized genomics: Perspectives from research participants. Public Health Genomics 15, 1 (2012), 22–30.
4. Pollack, A. 2013. F.D.A. orders genetic testing firm to stop selling DNA analysis service; http://www.nytimes.com/2013/11/26/business/fda-demands-a-halt-to-a-dna-test-kits-marketing.html?_r=0
5. 23andMe: https://www.23andme.com/
6. Personal Genome Project: http://www.personalgenomes.org/
Orit Shaer is the Clare Boothe Luce Assistant Professor of Computer Science and Media Arts and Sciences at Wellesley College. She directs the Wellesley College HCI Lab. Her research in HCI focuses on 3D, tangible, and tabletop interaction, as well as on computer-supported collaborative learning. firstname.lastname@example.org
Oded Nov is an associate professor at New York University’s Polytechnic School of Engineering. He received his Ph.D. from Cambridge University. His research focuses on behavioral, motivational, and dispositional aspects of social computing. email@example.com
©2014 ACM 1072-5220/14/09 $15.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2014 ACM, Inc.