Significant Science

Before we begin, Shirley, I’d like to give readers a bit of background as to who you are and why what you are doing is important. Those of us who follow the Open Science movement already know of your activity in FriendFeed on that topic and of your blog, of your doings on Twitter and of your status as an employee of the personal genetics company 23andMe
where you work as a scientist and contribute to its colorfully named blog, the Spittoon.

I asked to interview you because you embody so many interesting developments in so many different respects. You are a bona fide scientist, an employee at one of the most interesting startups in the world of Health 2.0, someone on the cutting edge of the commercialization of the fields of personal genomics and personalized medicine, and are well connected and respected in the world of Open Science (as evidenced by the fact that you have recently co-authored a paper, “Article-Level Metrics and the Evolution of Scientific Impact” with one of the leaders of Open Science, Cameron Neylon).

Thus, you are a fascinating case study of how scientists are helping Health 2.0 companies contribute to the generation, analysis and use of consumer-contributed data, which also makes you and your work of interest to those of us who are interested in the subjects of user-generated content in the health sciences and in the e-patient and participatory medicine movements, of which 23andMe is an exemplar. You also interest me because there are relatively few female leaders in the field of Open Science and I would like to use you as an example of female accomplishment in that realm so as to encourage young girls to get interested in the subject and in science in general.

I envision, then, as potential readers of this interview those who may be new to the subjects of Open Science, those who already follow the subjects of retail genomics, those who are potential users of 23andMe but who want more information about the science behind it, those who are interested in the subject of social networking in the field of consumer health and those who would benefit from a primer on the field of Web-based genetic testing services. Other potential readers could include physicians and other medical providers who may not realize that in coming years they may encounter patients appearing with material they have printed out from the sites of companies such as 23andMe and will ask their physicians how the physicians intend to use such information in the treatment of such patients—(given that the Web site of 23andMe includes such wording as, “With 23andMe, you and your doctor can determine whether certain medications will be right for you.”)

Let’s talk a bit about terminology. As I read around on the Web about 23andMe I came across various terms used in conjunction with it. Could you please tell us what each of these are, what the differences are between them and which ones accurately characterize 23andMe:

Personalized medicine
Personalized medicine is the use of information pertinent to the individual to guide healthcare rather than broad standards applied unilaterally over most patients. To some extent, medicine already uses some information like this — medical history, family history, clinical measurements — but to a large extent ignores other information that may be important, like genetic variants unique to that person. When people talk about personalized medicine, they often talk about incorporating the patient’s genetics into drug treatment decisions (should the patient take clopidogrel, or warfarin? Are they going to have an adverse response to choline ester drugs used as a muscle relaxant in certain surgeries?), an area known as pharmacogenetics (how a person’s genetics influences how they metabolize drugs). Specialties like oncology are already using these principles quite a bit. But personalized medicine is also becoming more about preventative medicine (knowing your genetic risks for health conditions so that you can take steps early to prevent or monitor for them).

Participatory medicine
I have less experience in this space, but from what I understand, participatory medicine describes a phenomenon where patients are more engaged in their healthcare and in the research related to health conditions they may have. The Society for Participatory Medicine says this on their website: “Participatory Medicine is a cooperative model of health care that encourages and expects active involvement by all connected parties (patients, caregivers, healthcare professionals, etc.) as integral to the full continuum of care. The ‘participatory’ concept may also be applied to fitness, nutrition, mental health, end-of-life care, and all issues broadly related to an individual’s health.”

Consumer genomics/genetics
This term describes genetic testing services — usually for health or genealogy — that are available directly to the consumer.

Gene-testing services
This probably describes any service that determines variants in one or more genes, usually for the purposes of disease screening. Examples are genetic tests for inherited diseases like cystic fibrosis or Tay-Sachs. The type of testing (whether it’s specific variants tested, a scan of the whole gene for mutations, etc; one gene or a panel of genes; etc) varies. Most of these are ordered through a professional healthcare setting, though some are also available to the consumer. These tend to be fairly expensive, at least compared to consumer genetics services that survey much more of the genome, but they are meant to test for a particular condition and so may be more comprehensive, sensitive, or specific with regards to that condition.

DNA tester
Not sure what this refers to… but perhaps it’s the actual test kit for genetic testing, e.g. spit tube or cheek swab set?

Retail genomics
I think this is just a press term for consumer genetics.

Personal genomics
I take this term to mean the application of genomic technologies to individuals as individuals, rather than genomics applied to anonymized individuals or populations, and where the individual has the ability to interact with and control their genomic data.

Personal Genome Service™ (which term 23andMe appears to have trademarked)
This describes 23andMe’s consumer services, which currently includes testing for over 500,000 genetic markers for the purposes of ancestry or health information. 23andMe also provides tools for interacting with and viewing the data, and educational and informational resources for exploring the intersection between your genetics and health-related topics.

Genetic testing
This encompasses any test that determines your DNA sequence at specific locations in the genome or the presence/absence of genetic markers or variants, usually applied in the context of health, forensics, or ancestry.

Could you differentiate for us what might be the differences between genetic testing and personal genomics? How does 23andMe, for example, differ from the firm Navigenics for instance, which uses the term, “personalized genetic testing?” Are 23andMe and Navigenics nearly identical or are their aims, services and target markets different? Sometimes, for example, the two firms collaborate in defending the value of their services as in their joint rebuttal to some of the statements made in an article in Nature (including some comments therein by Craig Venter), “23andMe & Navigenics’ Open Letter to Nature.”

I consider both 23andMe and Navigenics to be providing “personalized genetic testing”. Perhaps the reason why 23andMe uses the term “personal genetics” and Navigenics uses “personalized genetic testing” is that Navigenics puts less emphasis on the raw genetic data and exploring your genetics than 23andMe, and instead focuses on information on specific markers related to disease (23andMe also provides this). Navigenics currently takes a more physician-oriented approach, with on-staff medical personnel, while 23andMe currently focuses almost entirely on the general consumer. In the end, though, both companies are direct-to-consumer genomics companies that believe that individuals should have access to their genetic information, and so our interests are aligned in many policy discussions.

Could you, as a scientist, tell us a bit about that brouhaha and, in a way, doesn’t the fact that 23andMe and Navigenics have been a focus of discussion in such prestigious publications indicate that they have arrived as serious players in the field of genomics and are not simply trendy boutique services for health-obsessed, cash-rich consumers?

It’s true that there has been some public debate over personal genomics and DTC genetic testing. Similar debates and disruptions always happen whenever technology becomes readily available before social and regulatory frameworks have adapted to cope with them (e.g. peer to peer file sharing and the music industry). I think that most people are recognizing that consumer genomics is not something that is going to go away, and as two of the major players in this space, 23andMe and Navigenics have drawn much of the attention from those who would comment on this rapidly evolving industry.

Could you tell us a bit about your background? What was your graduate degree in and how did you come to work for 23andMe?

I came to 23andMe from a graduate program in Biomedical Informatics at Stanford University under the tutelage of Dr. Russ Altman, who is a leader in the field of pharmacogenomics and is also a member of 23andMe’s scientific advisory board. Being in his research group exposed me to the ideas of personalized medicine and the potential impact personal genetic information could have on healthcare. Bioinformatics students also gravitate towards rich data sets and emerging technologies — something that the personal genomics has in spades — so it’s no surprise that two of the founding R&D architects at 23andMe were also graduates of the Biomedical Informatics program. As I grew closer to finishing my degree, I knew that I wanted to work on something that had an impact on people and that combined my interests in science, society, and communication. Personal genomics is a natural fit because it touches on so many aspects of healthcare, education, culture, information, and technology. And the placement of 23andMe in Silicon Valley along with its accessible approach to genetics appealed to me.

Are you a bioinformatician—or are you a biomedical informatician? Is the latter an actual job title? What is the difference between the two fields?

Bioinformatics and medical informatics usually refer to biology-focused vs. medicine-focused informatics, though the distinction between the two is becoming less clear as translational science becomes more important. That is, there is more effort these days to connect discoveries made in basic research to treatments or practices we can apply in a medical setting (so-called “bench to bedside”). Biomedical informatics is a term Stanford uses to encompass all aspects of informatics in both of these areas. As far as job titles go, they so rarely describe what someone does accurately, so I won’t even go there.

Your official title at 23andMe is “Scientist, Content Curation,” correct? Could you tell us what you do on an average day? Do you help compile and edit the reports listed here, for example? What does that entail?

That is apparently my official job title, though like I mentioned before, it doesn’t nearly encompass everything I do. Similarly, it’s difficult to say what I do on an average day because there hasn’t really been an “average” day! I have helped on some of the reports listed on the 23andMe site, though most of those were already around before I started working here a few months ago. As we go forward, I’ll have a much more active role in finding ideas and information for new reports and implementing some of them on the actual site, as well as contributing regularly to the company blog and helping with the development of new phases of the online product.

For the actual health reports, our criteria for including a genetic association on our website is available in a company white paper, “Guidelines on Vetting Genetic Associations”, available here. Much of the job of curation is to determine whether literature gathered on a disease topic fits that criteria.

You provide a charming account of your first weeks on the job at 23andMe in the blog post, “New Job and Curation 101” and this passage indeed answers many of the questions I intended to ask you, “So what is personal genomics, anyway? We’ve known for a while that genetics – the sequence of DNA inside our cells – plays an important role in our form and functioning. Many diseases are caused by changes in DNA (often in genes, parts of DNA that code for proteins) that alter the normal functioning of cells, though not all genetic differences lead to negative changes. (Genetics can also tell us about ancestry – who is related to whom and the history of populations – but I won’t be addressing that in this post.) Where it gets personal is when you apply it to individuals, such as when someone gets a genetic test to determine whether they have or are at risk of developing or passing on a particular disease. Where it gets genomics is when we use high-throughput technologies to do what is essentially thousands of genetics tests at once. Put them together, and you get personal genomics.”

Is genomics, then, basically genetic testing on techno-steroids? What can scientists learn from genomics when it comes to specific diseases and what does that mean for sufferers of that disease? For example, I am quite interested in the subject of amyotrophic lateral sclerosis. In 2007 an article was published in the New England Journal of Medicine, “Whole-Genome Analysis of Sporadic Amyotrophic Lateral Sclerosis.”

Could you explain, using that example, what such studies mean for average people with the diseases so studied?

You could think of genomics as many genetic tests done in parallel — instead of determining the sequence at one or a limited set of locations in the genome, you determine it for many thousands or more. Scientists have been using genomic approaches to identify areas of the genome that have previously not been known to affect disease. This has led to the discovery of new genes involved in disease development, in some cases, though in many cases it’s still unclear why a particular region has a certain effect. They identify these regions by comparing the genomes of people with a certain condition (such as ALS) to people without that condition; regions that are significantly different between the two groups are more likely to have something to do with that condition. Right now, these studies don’t mean too much for average people with the disease — they simply add to our knowledge of genetic factors that may contribute to the disease. (In a few cases, the findings may eventually lead to a drug target or other therapeutic discovery.) They may, however, help identify people at greater risk for developing a disease.

You discuss single-nucleotide polymorphisms (SNPs) in your contributions to the SNPwatch Category on the Spittoon. Could you elaborate here, “It turns out that there are millions of single locations in the human genome where the exact sequence of the DNA might differ between two people, and these places, called single nucleotide polymorphisms, or SNPs, can contribute to differences we can observe, such as whether you flush when you drink alcohol or how easily you put on weight. 23andMe personal genomics kit determines what your sequence is for a representative subset of SNPs. Many are already known to be associated with certain conditions, and new research is being done every day to uncover more and more of these associations.”

To wit, how would I as a 23andMe user such information? As a one-time fact to bear in mind when I plan my daily meals or as a deadly serious matter requiring further investigation in consultation with a doctor or a genetic counselor?

There are different types of information you can get from your genetics related to health. Some diseases are known to be caused by specific mutations in genes, and so a positive result for one of these could mean a high probability that you have or will develop the disease (if it’s a dominant trait, you only need one copy of the mutation; if it’s recessive you need two copies) or could pass the mutation on to your children. 23andMe currently covers a couple dozen of these types of conditions. Your genetics also affects how you respond to certain drugs, and 23andMe reports on several of these as well. These two types of information are probably most important for a medical professional to know. For other conditions (most conditions), your genetics affects your risk to varying degrees, and environmental factors are very important. They are mostly things to keep in mind to be more informed.

Note that the other product 23andMe offers is Ancestry — and many people are using it to discover new distant relatives and explore their family trees.

What is a genomewide association analysis?

This is the analysis of a large number of genetic markers (typically in the hundreds of thousands, or more) over the entire genome to identify genetic markers that may be connected to a disease or condition. See two passages above.

I must say that you are a very skillful elucidator of these quite arcane matters. This passage is very helpful for example, “The genetic testing kit is one part of the product, but the other part is information – what knowledge is there about associations between the SNPs on our platform and health traits or conditions? What does your particular data mean? The science is far from exhausted on this subject, and in order to stay up to date with the research, 23andMe spends a lot of effort on curating the scientific literature for new genetic associations and presenting the information on our website for our customers.

Day to day, this means that we keep track of papers recently published in scientific journals, skim through to find ones that may have promising findings, and then vet these more thoroughly to see if they pass our stringent scientific standards. If they do, we extract the bits of information we need and put the bits together in reports that will eventually become part of the content on the website. It’s a job that definitely benefits from an organized system and an eye for detail – as well as a sense of curiosity.”

That is a very helpful passage and I would say that 23andMe excels at public education and building bridges between the general public, the increasingly large numbers of sophisticated health information seekers known as e-Patients (as exemplified by e-Patient Dave) and that the Spittoon and the blog of PatientsLikeMe, The Value of Openness are quite fascinating examples of a new genre—the Health 2.0 company public education blog. I am hoping that medical librarians read this interview, given that more and more consumers are going to learn about developments in clinical research and treatments from such blogs and we need to know about such forums.

And on the related matter of online patient communities and patient and consumer health social networking could you discuss 23andMe’s hopes of creating an active community component? Is that going well? Are you getting much traffic in your online communities? Garnering actual traffic seems to a challenge for many consumer health sites—has 23andMe had much luck in that respect?

23andMe has an active, thriving user community, roughly divided into those interested in Health and those interested in Ancestry, though all users can post in all areas of the community forums. As with any community, we have many who are very passionate about their interests, and many who are simply curious about learning more or in finding people like them. We also have surveys which users are encouraged to fill out about various health and environmental topics, for the purposes of conducting our own research studies. Our first paper using data collected in this way was recently presented at the American Society for Human Genetics conference, with novel associations presented for hair curl, asparagus anosmia (the ability to smell asparagus metabolites in your urine), and hair color.

There have been some layoffs at 23andMe in recent weeks but influential journalists like Wired Magazine executive editor Thomas Goetz seem enamored of 23andMe’s services. That is a pretty strong endorsement. Are you confident about the long-term outlook for 23andMe? I heard your colleague Alex Khomenko speak at the e-Patient Connections Conference 2009
and he certainly did a good job in conveying his conviction that 23andMe is capable of serving the public good by doing good science in the commercial sector. Could you please tell us at what other conferences representatives of 23andMe will appear?

I don’t have any information about conferences in 2010, though there is interest in organizing a session at next year’s ASHG meeting in Washington, D.C.

One of the things that I like about 23andMe and me is that it allows you great freedom to keep up your valuable role in the Open Science community, to the great benefit of science, and therefore of patients down the pipeline. Your elucidation on your blog of such complex matters as the basics of scientific curation is really engaging and edifying, “In science, curation involves organization of scientific knowledge and data. An area where this has been especially important is the life sciences, as the amount of information being generated by high-throughput experiments, large-scale projects, and scholarly publishing has skyrocketed. In order to manage this information and render it useful to others, the field of biocuration was born. Any database that organizes scientific knowledge – UniProt (the Universal Protein resource), FlyBase (database for that very important model organism, Drosophila), PharmGKB (a database focused on how genes and drugs interact), etc – depends on curators to keep the information up to date and easy to use.

And so it is with 23andMe.”

And on that topic, I’d like to get back to your reputation as a go-to person in Open Science. For example, I am helping to plan a conference on the subject and went to one of the giants in the field, Jean-Claude Bradley for advice on determining the possible attendance at such a conference. He suggested that I consult you and you very helpfully suggested that I send up trial balloons in Twitter and FriendFeed. The result was that there did indeed seem to be some excitement about such a conference. I would like to ask you, then, how you use Twitter and FriendFeed to keep up on developments in Open Science and science generally. And does 23andMe have a specific Twitter strategy?

I personally use Twitter and FriendFeed mostly to keep a finger on the pulse of science and technology news and developments. I haven’t contributed as much as I’d like to but it’s very useful for me just to stay abreast of everything that’s happening in the world.

23andMe has a Twitter account as you show above. It’s used mostly to communicate with people interested in our company and our service, to help spread stories people are sharing about their experiences with our products, and to broadcast news about the company or the industry.

You say on your blog, “Since my work is even more directly tied to the literature than it was as a graduate student in academia, I’m also developing an enhanced awareness of issues surrounding scientific publishing – those related to standardization and metadata, publication bias towards positive results, and closed vs. open access.” Could you discuss some of your comments in the article you wrote with Cameron Neylon, “Article-Level Metrics and the Evolution of Scientific Impact?” I was especially interested in the section, “The Trouble with Comments” and found this passage quite interesting, “A spirited, intelligent comment thread can also help raise the profile of an article and engage the broader community in a conversation about the science.

Unfortunately, commenting in the scientific community simply hasn’t worked, at least not generally…Part of this resistance to commenting may relate to technical issues, but the main reason is likely social. For one thing, researchers are unsure how to behave in this new space. We are used to criticizing articles in the privacy of offices and local journal clubs, not in a public, archived forum. Junior researchers may be concerned about the potential repercussions on their own careers…

Another issue is that the majority of people making hiring and granting decisions do not consider commenting a valuable contribution.”

As a fairly recent graduate in the sciences, could you please elaborate on the matter of commenting? Do you think that one can enhance one’s reputation by blogging but that commenting is much less fruitful in that regard? Your comments here are well taken, “…if there is no reward for quality contribution then people will struggle to justify the time involved in generating high quality comments.”

Commenting is one part of the spectrum of scientific content contribution. At one far end is formal publication in a peer-reviewed journal. Commenting online is near the other end of the spectrum in terms of perceived value, but its potential impact is fairly large. Just as reviewer comments on a manuscript can be very helpful, so can less formal comments on a paper. But the barrier to entry is so much lower that commenting has the opportunity to have that much more impact. Of course, that low barrier also means potentially more noise to filter out. And the lack of reward for commenting means that there is little incentive to contribute good comments other than goodwill towards fellow scientists or the scientific enterprise. There is also a tough balance between anonymity and comment quality — anonymity may discourage accountability and foster unhelpful comments, but lack of anonymity may result in reluctance to post honest criticisms, or anything at all.

I do think that blogging is currently placed in higher regard than commenting – depending on the subject matter and quality of the blog, of course. It is also easier to compile a record of blogging than of commenting, since a blog is an archived set of posts in one place, whereas comments may be scattered across a multitude of sites. But Faculty of 1000 is a respected site that is in essence a collection of comments on specific papers, and so there is hope for expanding the scope of scientific contribution to include commenting more generally.

As someone who spends a lot of time looking for grants to list on the site I work on Scangrants, I was quite interested in your comments that those who make granting decisions do not consider commenting a valuable contribution. It will be interesting to see if that changes as those who make grants, such as private foundations and scientific societies, start to adopt social media tools in order to showcase the benefits of their grantmaking activities and engage public interest in them and if the number of venues for comments on scientific projects will increase.

I just left a comment on your article and doing so was fairly easy (provided that one remembers the password for each site where one wants to leave comments). But as you point out, it is hard to see the incentive for doing so except for the warm fuzzies one gets for encouraging scholars engaged in producing such edifying articles.

Could you tell us about the process of writing for PLoS Biology? How did you and Neylon decide to publish your article there? Whose idea was it? Is publishing in such Open Access periodicals any different vis-à-vis process and reaction from readers from publishing in toll access ones?

Writing the ALM article for PLoS was a relatively smooth experience. Cameron had been approached by the PLoS team to write the article, and he contacted me to co-author with him given that it would echo many of the issues I brought up in a recent blog post I’d written. The actual process of submitting the article was done through one of the PLoS Biology editors, rather than the online submission system, and I have not actually gone through the process at a toll-access journal before (I’ve submitted to BMC journals in the past), so I can’t comment too much on the differences. The reaction to the article has been great so far — several thousand views, quite a few comments and ratings, etc. Publishing in an Open Access journal like PLoS (without added media involvement) I’m sure raised the profile of the article to many more viewers than would be possible in a closed-access journal, and being able to gauge this reaction through ALMs is, of course, unique to PLoS.

Finally, who are your heroes in science, scholarship, technology and in any other realm?

Most of the people I admire and look up to are people I’ve had the pleasure of meeting or working with in some capacity. My graduate advisor, Russ Altman, is certainly one. But there are too many to list exhaustively so suffice it to say that these are people who are great teachers, great examples, who are passionate about things and work hard to make things happen, who somehow are able to keep up with everything that’s going on and help edify the rest of us through blogs and social media and podcasts, etc. If you follow the FriendFeed Life Scientists room, you will shortly figure out many of the people I’m talking about.

Thank you for your time, Shirley.