“Let’s pull these technologies out of the ivory tower”: The politics, ethos, and ironies of participant-driven genomic research

Calculable selves, predictable futures

One of the most powerful cultural shifts accompanying the rise in PDGR relates to a rise in social expectations that individuals should monitor and manage their own health. This expectation undergirds terms such as “participatory medicine,” “Health 2.0,” “Medicine 2.0,” or “Health 2050” (Harris et al, 2013; Hughes et al, 2008; Kaye et al, 2012; Swan, 2012). The communities that have devised and promoted these terms are committed to the idea that people are no longer ‘patients’ but are instead ‘participants’ and instigators of their own health management plans. More precisely, the expectation is that individuals participate proactively in health monitoring processes that enable them to preemptively identify and manage disease risks or factors and to maintain or enhance health and wellbeing (Juengst et al, 2012). These might be accomplished, for example, through the optimization of sleep, diet, physical activity, and cognitive performance.

Technology is key to this optimization, including an array of products such as low-cost smartphones, wearable gadgets, biosensors, and cloud-based services that track, store, and aggregate data about an individual’s genome, microbiome, biophysical functioning, and physical environment. Data derived via these technologies are not only quantitative but also qualitative. The aforementioned metrics can be coupled with self-reported data on one’s symptoms, mood, sociability, or stress. Self-tracking thus leads to the collection of information that is ‘actionable’ in that it enables feedback loops for behavior changes that are intended to foster better or different outcomes (Swan, 2013).

Genomic research beyond the traditional laboratory

The expectation that individuals track health-related information and manage their health – and be empowered to do so – has also led to changes in how genomic research is conducted (Wyatt et al, 2013). Perhaps the most important of these changes has been to harness Health 2.0 technological and cultural developments in attempts to overcome the methodological challenges of underpowered and proprietary genomic research studies. PDGR groups have a heavy online presence and capitalize on the interactive nature of the Internet to gather and present genetic data to large communities of web-based research participants and consumers of direct-to-consumer (DTC) genomic technologies (MacArthur, 2009). This research paradigm not only improves statistical power for studies of genetic risk variants and environmental risk factors, but it also means that genomic research increasingly takes place outside the bounds of the traditional research laboratory.

Citizen genomics

The activities of these organizations frequently reflect the ethos and practices tied to the growing movement of ‘citizen science’ and public participation in scientific knowledge production (Lengwiler, 2008; Kelty and Panofsky, 2014). The citizen science movement aspires to open the sphere of scientific expertise by drawing on the energy and capacities of large and widely distributed communities and by involving the participation of ‘lay scientists’ in research design and funding, and in collecting, analyzing, applying, and disseminating data (Irwin, 1995; Prainsack, 2013). This activity has largely been driven by grassroots organizations and community-based organizing related to environmental (in)justice, and has involved attempts to collect data to demonstrate the disproportionate and damaging effects of environmental exposures of some groups and neighborhoods (e.g. Allen, 2003; Brown and Mikkelsen, 1997; Brown et al, 2011; Fishman, 2000; Ottinger, 2013). In recent years, the term ‘citizen science’ has also been used in other contexts to describe web-based science projects that use ‘lay citizens’ as data collectors – but without its earlier political dimensions or commitments. Exemplars of contemporary citizen science projects relate to large-scale ornithology, astronomy, and conservation (Citizen Science Alliance, 2014; Trumbull et al, 2000). Recent biomedical projects within and outside of traditional research contexts have also been advanced through citizen science strategies.

These projects and the organizations that have propelled them reflect the ethos of the open-science movement, which rejects the corporate privatization and patenting of scientific data, knowledge, and biological processes in favor of cooperative research and the free sharing of data and knowledge (Delfanti, 2011; Kelty and Panofsky, 2014; Vayena et al, 2015). PDGR groups have actively taken up the ethos of citizen science and ‘democratic science,’ with many online communities employing open-source computer programming, collaborative learning, and the language of empowerment (especially regarding the rights of participants to access, share, and own their data) to expand the notion of scientific ‘expertise’ and participation in research beyond universities, hospitals, and companies.

Crowdsourcing – a method borne out of the Internet’s capacity to recruit large numbers of people – allows for the accelerated and inclusive gathering of data. Wikipedia is the most well-known example of this type of data collection. It is based on a belief about the ‘wisdom of crowds’: the power of numbers will lead to greater accuracy and more complete information (Surowiecki, 2005). The perspective is that when all people are permitted to participate, the resulting information is at once richer and more accurate. This method has been adopted in biomedical research, for example, through the web-based submission of biological data gathered through kits or questionnaire data on genetic, phenotypic, or lifestyle matters. Within the arena of Health 2.0, the most prominent example of this thinking is Patients Like Me. Patients Like Me began as an online repository for ALS patients and their caregivers to document the effectiveness and side effects of medications. However, this soon grew to a broad-based website for ‘patients’ to report their symptoms and symptom relief from a myriad of health conditions, ranging from fibromyalgia and depression to epilepsy and Parkinson’s Disease. Anyone (with or without a medical diagnosis) is welcome to participate by adding her own personal information. These resulting data have already been used in publications (e.g., Wicks et al, 2011) and are available for use by pharmaceutical companies and other researchers who partner with Patients Like Me via individual contracts.

Crowdsourcing has facilitated large-scale data collection for a wide range of PDGR-labeled projects and organizations. The logic of harnessing wisdom, talent, and free labor through the Internet is very appealing as it has created new opportunities for businesses and researchers, and for engaging people with and without formal expertise (Howe, 2006). For example, DIYgenomics, an organization for crowdsourced health research, aims to increase participant-led collaborative research in order to build large biobanks whose information can accelerate advances in health management and preventive medicine (DIYgenomics, 2015b). Users are invited to upload individual data gathered from smartphone applications to contribute to projects ranging from studies of vitamin deficiencies to telomerase activation and social intelligence (DIYgenomics, 2015a). In capitalizing on social media technology and the growing imperative to share information about oneself with online communities, PDGR is constructing new niches for active biological citizenship, a means of finding political, social, and other affinities with others on the basis of shared biological characteristics (e.g., genetic, disease status) (Levina, 2010; Rose and Novas, 2005).

By challenging professionally mediated access to genomic information, PDGR resembles the wave of commercialized DTC genomic testing companies, such as 23andMe, that disrupted traditional clinical gatekeeping on the risk assessment side of PGM. Customers of 23andMe can directly order test kits online at increasingly low cost, provide saliva samples, and then gain online access to a set of personalized disease risks, phenotypic traits, and ancestry-related data based on a scan of around 500,000 single-nucleotide polymorphisms across the genome (McGowan et al, 2010). A research arm of 23andMe calls for customers to contribute their data to larger research studies. It has been a forerunner in maximizing the capabilities of crowdsourcing, a key tool in the emergent field of PDGR. Yet, we did not include this company in our sample, as consumer–participants have a low-degree of control over agenda setting and resource control (Kelty and Panofsky, 2014) – it is not itself ‘participant-driven,’ but a for-profit company with ‘participatory’ dimensions.

Present study

This study draws on interviews with a diverse array of organizations and individuals that contribute to the expanding ecosystem of ‘PDGR’ – including academic groups, start-ups, activists, hobbyists, and hackers – in order to compare and contrast how they relate their stated objectives, practices, and political and moral stances to institutions of expert scientific knowledge production.

This study developed as part of our ongoing research on the emergence of personalized genomic medicine, specifically, on early users’ motivations for engaging with DTC personal genome testing, a technology born of genomic research, the Internet age, and commercial entrepreneurialism (McGowan et al, 2010). Angrist (2009) refers to early adopters of this technology as utilizing “citizen science,” and Prainsack (2014, p. 155) has documented how the leading commercial provider of personal genome testing “fits into the bigger picture of citizen science.” We have argued elsewhere that DTC personal genomic testing rearticulated the relationship between biomedical and information technologies through its co-constitution by developers and users (McGowan et al, 2010), and this led us to question whether and how PDGR rearticulates the relationship between biomedical research and information technologies. This nascent intersection of participant-driven citizen science and genomics led us to further explore this phenomenon among non-commercial entities in order to assess the goals, values, practices, and future visions among organizational leaders involved in PDGR, with a particular focus on notions of knowledge producers, research questions and settings, research funding, and research ethics.

Kelty and Panofsky (2014, p. 2) have argued that “traditionally, scientists and medical researchers have preferred research subjects to remain ‘subject’ to their direction and control, not ‘subjects’ in the sense of people who ‘talk back’ to experts and insist on representing or pursuing their own interests.” We use the term ‘participant-driven genomic research’ purposefully, as an umbrella term to describe commitments that each of the organizations we interviewed share that challenge the traditional scientist/subject relationship: involving non-traditional ‘participants’ (e.g., ‘lay’ people, consumers, patients, scientists in non-academic or commercial settings) in the multidimensional aspects of genomic research, including design, funding, conduct, analysis, and/or distribution of findings. While this is not necessarily a term that each organization would ascribe to themselves, it adequately captures the organizational features that we will be highlighting in our analysis.

Sampling strategy

We first learned of the existence of PDGR organizations through conducting research with early users of DTC personal genome testing (e.g., DIY Genomics and SNPedia). The aims and approach of this earlier project provided us with a census of stakeholders, including a growing community of enthusiasts pursuing what we call ‘participant-driven genomic research’ (PDGR). Through this research we were became aware of groups like bioCURIOUS, Genomes Unzipped, and Personal Genome Project. These interviews sensitized us to a new development within genomic research: the emergence of groups like these that exist outside of mainstream academia and conventional genomic companies. We did not initially label these groups, but sought to identify groups that expanded our sample to include other organizations conducting genomic ‘research’ very broadly construed in an effort to recognize the broader array of emergent groups concerned and invested in public participation in research to inform personalized medicine. This approach follows from our methodological framework of Social Worlds and Arenas Theory (Clarke, 2005; Clarke and Star, 2007), which requires that we map and ascertain the perspectives of all actors within a social arena that has a ‘going concern’ with the phenomenon in question (Hughes, 1971; see also Fujimura, 1996; Clarke, 1998).

In addition to sampling leaders of organizations we encountered in the course of our empirical research, we also conducted a web-based literature review using Google Scholar and PubMed, combining the search terms “online,” “genomics,” and “citizen science” to assess the online presence and activity of PDGR. Through this approach we identified two additional organizations: DIYbio and My Daughter’s DNA.

We do not mean to imply that there is uniformity or even cohesion across these groups: they vary considerably in terms of their approaches, activities, organizational structures, and dimensions of participation (Kelty and Panofsky, 2014). We sampled, interviewed, and analyzed these organizations together because they each explicitly identified themselves as engaging a range of actors that would not typically be considered experts in scientific knowledge production in the design, conduct, and distribution of genomic research and attendant technologies.

Recruitment

We contacted the executive directors, founders, or active spokespeople from each organization with a personalized email or phone call in order to describe our research interest in their organization. Recruitment occurred between April and September 2013. We contacted leaders from 23 organizations, and successfully recruited key informants from 12 organizations. We interviewed a single informant from seven organizations, two informants each from four organizations, and three informants from one organization, for a total of 17 interviews with 18 individuals. Informed consent was secured in writing and verbally at the start of each interview, at which point interviewees also indicated whether they wished to have their name associated with their responses. SC and MM conducted the interviews via Skype or by phone using a semi-structured interview guide developed by the research team. The interview guide contained questions pertaining to the role of the interviewee in the organization; the goals, values, funding, and research priorities of their organization; beliefs about impacts of PDGR on healthcare and medical practice; ethical, legal, and social challenges of PDGR; and their visions of the future of the field. Questions were asked flexibly so that interviewers could respond appropriately, probe answers as necessary, and collect consistent information across participants. Interviews lasted 45–75 min and were recorded using Audacity 2.0.3 software.

Data analysis

Following standard strategies in the social sciences, interviews were transcribed verbatim, reviewed for quality control, and coded using a codebook with precise definitions for thematic analysis. In order to ensure reliability, the research team first coded a batch of initial interviews together to develop the codebook and establish guidelines for applying codes. The research assistants charged with coding all of the interviews then coded each interview using Atlas.ti 6 qualitative data analysis software. The research team drafted summaries of coded data, working across summaries to identify major themes (Merriam, 2009; Braun and Clarke, 2006). In this paper, we rely on thematic analysis of codes that captured the goals and ethos of the various groups, descriptions of their methods and practices, and asserted uses of technology. We used these codes to characterize the organizations involved in PDGR and to distinguish their particular objectives vis-à-vis mainstream genomic research, moral and political principles, and any tensions therein.

Rare disease patient advocacy

PDGR has increased patient advocacy on the web related to rare diseases (Kelty and Panofsky, 2014). A prominent example is MyDaughtersDNA.org, led by Hugh Reinhoff, who developed a website as he sequenced a portion of his daughter’s DNA to investigate the genetic basis of her undiagnosed condition. In our interview, Reinhoff noted that his perception of unmet needs in the rare disease research sector and medical system drove him to establish his own website to conduct research on his daughter’s undiagnosed condition. He explained his efforts: “If we really wanted to know, I had to spearhead the effort myself, and the first step in that was really getting a thorough inventory of her chronicle phenotype.” Reinhoff characterized his site as a forum for altruists and social activists aiming to increase awareness and raise funds for understudied genetic diseases. He has been lauded by others in the patient advocacy world as a trailblazer in inspiring patients and parents of children with rare diseases to create and participate in online communities in order to discuss and investigate complicated diagnoses (Maher, 2013).

Efforts like My Daughter’s DNA benefit from the availability and popularization of web-based social networking platforms and crowdsourcing to engage patients and their families in collecting and analyzing data (Vayena and Tasioulas, 2013b). Because rare disease communities have historically been marginalized by academic, commercial, and federally funded science (Novas, 2015), PDGR aims to develop new health measures, disseminate research to large numbers of participants, and generate results faster than the time frames associated with traditional approaches to genomic research.

However, rare disease patient advocacy, especially in the genetics arena, pre-existed this most recent incarnation. Throughout the 1980s and 1990s, groups emerged to provide support for those with rare diseases (Rabeharisoa and Callon, 1998). During this time, patient groups’ involvement with disease research went beyond advocacy to develop ‘partnerships’ with scientists (Rabeharisoa, 2003). As has been detailed elsewhere, patient groups have turned to ‘evidence-based’ activism to strengthen their cause, developing their own networks of knowledge production that become indispensable to scientists as a way to become ‘insiders’ in scientific research (Rabeharisoa et al, 2014a). Genetic Alliance is the paradigmatic and most prominent example of these types of advocacy organizations. With time, Genetic Alliance also began to advocate for genetic research and, moreover, to place the findings and potential benefits of research directly in the hands of affected individuals (rather than through traditional healthcare, academic, or research-based organizations) (Lambertson and Terry, 2014). Arguably, organizations like Genetic Alliance have led the way for other disease-based and data-driven PDGR movements in the United States.

Democratizing access to genomic information

Hathaway’s quote is particularly salient in the realm of rare disease advocacy, which intends not only to hasten the pace of traditional genomic research, but it is also committed to challenging the research regulation and clinical trial paradigm that demands that research generate generalizable knowledge (and therefore applicable to wide populations) rather than personally relevant knowledge (and applicable to very narrow populations).

One point that is pertinent to rare disease advocacy is that research is characterized as underfunded through traditional research mechanisms. This makes more pressing for them the need to take data – and moreover funding – into their own hands. Hugh Reinhoff’s self-funded My Daughter’s DNA drew upon strong professional connections to the commercial genomic research industry to help him sequence his daughter’s DNA outside of the academic and commercial sectors, epitomizing Kelty’s (2010) figure of the Victorian gentleman scientist with the means and connections to conduct amateur research in his salon. While clearly pulling on the same purse strings as traditional research alliances, My Daughter’s DNA demonstrates the still-requisite need for significant social and market capital.

Effectively, the site additionally functions to fulfill the founders’ goals of meeting the translational imperative of genomic research, providing a direct link between institutional science and lay concerns about genomics and fostering the translation of genomic research to the public. However, despite rejecting private ownership of genomic information, SNPedia is not completely disentangled from corporate interests. This group purchases inexpensive server space from Amazon.com to store data uploaded to SNPedia, which raises questions about the implications of shift from private ownership to private storage of genomic data. Furthermore, using the site also requires a certain amount of genetic literacy and knowledge that SNPedia itself does not provide.

Deinstitutionalizing science: Practices, education, and outreach

Groups organized around self-tracking, self-experimentation, and participant-driven approaches in genomics are not solely, or even primarily, focused on concerns about personally relevant hereditary disorders. Self-ascribed as ‘health hackers,’ ‘biohackers,’ or ‘biopunk’ hobbyists, these groups constitute movements of people organized as social networks, and include latter day Victorian amateur scientists (often graduate students and hobbyists), artists, hackers, and software specialists who are guided by a set of goals that are compatible with but an extension of ‘citizen science’ movements and principles (Keulartz and van den Belt, 2016).

Their engagement in genomic research, both online and in physical spaces outside of traditional research institutions, is driven by the wish to be liberated from traditional relationships between research, universities, and the market. They seek to redefine genomic ‘expertise’ and the ownership of data, and to conduct research in deinstitutionalized spaces and develop or access low-cost and shared apparatus and laboratory resources and software (freeware). Their goals are manifold: to engage the public in science, to promote the use of open-source software, to transpose hacking practices into the realm of genomics and synthetic biology, to de-institutionalize research, and to develop innovative solutions to scientific problems or simply matters of curiosity (Delfanti, 2012).

One of the groups mentioned by Gentry – DIYbio – was intentionally organized (in 2008) to become a home for ‘outlaw’ biologists functioning primarily outside the academy and industry and working to demystify scientific systems and protocols (Kelty, 2010). DIYbio (whose logo is a fist with a pipette) (www.diybio.org) embodies the ‘biopunk’ spirit in that it uses community spaces – from garages to kitchens, closets, and schools – to provide lay biologists and engineers with the know-how and tools to conduct genomic research cheaply, investigate their own questions, share data with peers, and form communities around the values of openness in knowledge production (Ledford, 2010a, b; Keulartz and van den Belt, 2016).

We see here the metaphorical and literal use of “openness,” where the open lab space design and shared ownership of equipment reflects the underlying ideological commitments to deinstitutionalizing science – a commitment to doing science differently once it is wrested away from academic and other institutions and put in the hands of a community. Whether this architecture is largely symbolic or materially transformative is an open question, but this was a conscious decision to disrupt materially conventional ways of ‘doing science.’

These practices have elicited anxiety among governments and regulatory bodies (e.g., FBI and the Presidential Commission on Bioethics) due to concerns about biosecurity and safety (Bennett et al, 2009; Whalen, 2009). However, these groups countered with the argument that their practices foster much-needed public education on genetics and scientific citizenship (Delfanti, 2012). In our interviews, spokespeople for DIYbio describe their spaces as hubs to engage lay people in the skills and familiarity to understand genomic science, debate its implications, and equip them with the tools to participate in everyday decision-making in our technoscientific societies. For example, the group in Los Angeles offers courses on basics of polymerase chain reactions (PCR), teaching participants to sequence their own DNA from cheek swabs and discover predispositions to illnesses; the New York City group, called Genspace (2014), has a “PCR and Pizza” night and “Biohacker Bootcamp,” and the Brooklyn group has a storytelling evening (a cross between biotech and sci-fi class).

Increasing affordability of technology

Garvey’s eagerness to help represents his and others’ belief that wresting genetic technologies away from traditional manufacturing, and moreover, commercial patenting, will revolutionize global access to genetic testing in a way that obviates the need for traditional experts or traditional labs. While some may see this as a naïve reading of global science and global marketing, there is no reason to doubt his sincerity in wanting to make this happen.

Openpcr.org illustrates what Delfanti (2012) characterized as a notable contradiction in the values of many of these groups. In this case, on one hand, the ‘hack’ was to undercut PCR machines on the market, and to make them open source and relatively cheap. On the other hand, the very biotechnology companies and research institutions that members of openpcr.org initially aimed to subvert are now buying their low-cost PCR kits. The politics of these groups are therefore often paradoxical: they aim to undermine intellectual property rights and monopolies of capital and power in biotechnology, yet these organizations are often simultaneously eager to establish small start-ups which will compete with the big biotechnology companies, and maybe even become one of them. Macauley characterized the receptiveness of bioCURIOUS – a Silicon Valley-based biotech hackerspace – to such eventualities as reflective of its open-science ethos: “We care to be fairly radically open science. People come in, and if somebody wants to work on something and commercialize it, they’re free to do that.”

The spirit of bioCURIOUS, DIYbio, openpcr, SNPedia, and biohackers is therefore both subversive and entrepreneurial: biohackers draw from the mythologies associated with the start-up success of Silicon Valley – Google, Apple, etc. – while simultaneously critiquing the hierarchy and profit incentives of universities and corporations (Delfanti, 2012; Wohlsen, 2008). Like their Silicon Valley heroes, they embrace open information sharing, non-competitive collaboration, and market-deflating ‘hacking,’ but only until there is money to made. At that point, their libertarian spirit equally easily justifies joining the market to capitalize on their work for financial profit. In this way, their political economy operates very much like traditional genome science, where data-pooling and the free flow of ideas is endorsed right up until the intellectual property involved begins to show signs of scientific value.

Re-envisioning research participation and funding

While many of the PDGR groups discussed thus far explicitly position themselves outside of mainstream scientific channels and in opposition to the elitism of university research departments, other groups operate in partnership or as part of academic science. These groups largely engage PDGR through re-imagining what it means to be a research participant in an academic setting, and challenging traditional research funding paradigms.

The project, whose first ten participants (“PGP-10”) famously donated samples for analysis between 2006 and 2007 (Harmon, 2008), has been influential in guiding debates about the ethical and legal aspects of sharing genomic information. In particular, the PGP has challenged guidelines and regulations governing human subjects research that have held dear the exceptionalism of personal genomic information and importance of individual genetic privacy. To enroll, participants must provide open consent – foregoing privacy protections in favor of full disclosure of personal genomic information – which makes it impossible to guarantee participant anonymity because of the potential to re-identify individuals through their DNA sequence (Angrist, 2009). Interestingly, several students and researchers who were associated with Church’s Harvard laboratory have gone on to become active figures in the domain of DIYbio.

This is both a statement about how PDGR challenges research ethics and policy, as well as aspirations for how research epistemologies may be reshaped through participant-driven research efforts. Our informants did not view the re-imagining of the researcher–participant relationship as novel. Rather, it was framed as a necessary step in addressing previously unanswerable and ‘unfundable’ research questions.

A significant need and subsequent benefit of participant-driven research is the availability of large volumes of ‘big data.’ Academic research groups have also taken advantage of this method to tap anonymous participants’ ‘brain power’ through online games (Good and Su, 2011). For example, researchers at the Departments of Biochemistry and Game Science at the University of Washington developed an online protein folding game in 2008 (Foldit, 2014; Khatib et al, 2011). With over 240,000 registered players, participants’ pattern recognition and puzzle solving skills provide steps towards solutions for the design of protein structures relevant to drug development for diseases including HIV, Alzheimer’s, and cancer.

These hybrid practices are built on new interdependencies between research departments, private investors, and individuals – and they bring new regulatory questions related to research practice, credit-sharing, and ownership over products and findings, and the need for researchers to be both calculating and entrepreneurial in framing their research approaches (Rajan, 2006).