23andMe
From Razib: "Creative destruction in the personal genomics industry?"
I’m hearing about rumblings at 23andMe, and not in a good way.
Last year: "23andMe co-founder Linda Avey leaves."


From Razib: "Creative destruction in the personal genomics industry?"
I’m hearing about rumblings at 23andMe, and not in a good way.
Last year: "23andMe co-founder Linda Avey leaves."
A couple of weeks ago, the Texas Tribune reported on an investigation of the archiving of blood samples taken from newborn infants: "DNA Deception".
For decades, the state has screened newborns for a variety of birth defects, pricking their heels and collecting five drops of blood on a paper card. Until 2002, the cards were thrown out after a short storage period. But starting that year, the state health department began storing blood spots indefinitely, for “research into causes of selected diseases.” Four years later, DSHS began contracting with Texas A&M University’s School of Rural Public Health to warehouse the cards, which were accumulating at a rate of 800,000 a year. State health officials never notified parents of the changes; they didn’t need consent for the birth-defect screening, so they didn’t ask for it for research purposes. The agency’s rationale was that it let parents who asked opt out of the newborn blood screening and de-identified all of the samples before shipping them off (emphasis added).
So much for informed consent. "We're from the government, and we're here to help you."
The state was sued by parents last year and rapidly settled the lawsuit before pre-trial discovery. Now, it is suspected that the state was trying to avoid drawing attention to some of the uses of the blood samples -- including several hundred which were used to develop a forensics database of mtDNA variants.
E-mails indicate that in 2003, when the agency started to release blood spots for outside research, officials knew they had a parental consent issue on their hands — but tried to avoid it. When a researcher proposed a project, the director of birth defects monitoring wrote that he’d “prefer to not have to go through” the process of getting consent. Another agency official responded that parents "never consented for blood spots to be used for research. … On the other hand, I believe [the health department] already uses (deidentified?) blood spots for some research, so that might not be a big deal.”
All states now test for metabolic disorders in newborns; the tests require only a blood spot on treated paper. The National Newborn Screening and Genetics Resource Center has more information on the specific tests and a very up-to-date list by state. It is amazing to me, as someone who has had four kids in the last ten years, just how quickly these screening programs became universal.
It is therefore hard for me to believe that Texas is going to be an exception. Surely we'll discover that some other states are archiving these blood samples instead of destroying them? Checking them out to researchers for no-consent research?
Daniel MacArthur reports from the Advances in Genome Biology and Technology meetings are full of little snippets of next-generation sequencing news; good if you're interested but don't follow the developments closely: "Belated news from AGBT", "Pacific Biosciences introduces new third-generation sequencing instrument at AGBT".
UPDATE (2010-02-28): And "New players in sequencing debut at AGBT".
If that's true, and if it can be done at scale, it is extraordinarily cool: reads of unlimited length would profoundly transform genomics
Amy Harmon reappears in the NY Times science page this week, with a series on the clinical trials of a targeted cancer drug ("A Roller Coaster Chase for a Cure").
Dr. Flaherty, who has a near-photographic memory, was not accustomed to rereading. But in his campus office that morning, he scrolled through the article on his computer again to be sure he had understood. The presence of the same B-RAF mutation in so many cancers, he thought, meant it was one of the biggest genetic smoking guns yet identified in cancer. A drug that blocked the protein made by the defective gene might have enormous consequences for patients — and he knew of one that just might work.
This is where the "rubber" of personalized medicine "hits the road", so to speak -- if we can find drugs that treat the specific mutations that cause a person's cancer, then there may be hope in other kinds of interventions targeted to a particular genotype.
The Oscillator's Christina Agapakis reviews some work in synthetic biology -- "Expanding the genetic code"
But what if instead of mutating individual tRNAs, you could make a whole parallel genetic code in a living cell? An awesome paper in this week's Nature makes progress towards this goal, by using directed evolution to design a ribosome that reads four letter codons instead of the normal three. With a four letter code, you could potentially program 256 different amino acids, to create altered proteins or entirely different biological polymers.
This seems like the kind of thing that ought to be encouraged. You know, so synthetic organisms won't be able to eat us so easily. Of course, that's what they thought about engineering lysine-deficient dinosaurs in Jurassic Park.
A couple of long stories in the New York Times Magazine this weekend caught my interest. One of them covers the emerging world of university competitions in synthetic biology:
Over the past five years, iGEM teams have been collaboratively amassing a centralized, open-source genetic library of more than 5,000 BioBricks, called the Registry of Standard Biological Parts. Each year teams use these pieces of DNA to build their projects and also contribute new BioBricks as needed. BioBricks in the registry range from those that kill cells to one that makes cells smell like bananas. The composition and function of each DNA fragment is cataloged in an online wiki, which iGEM’s director calls “the Williams-Sonoma catalog of synthetic biology.” Copies of the actual DNA are stored in a freezer at M.I.T., and BioBricks are mailed to teams as red smudges of dehydrated DNA. Endy showed me a set stuck to paper, like candy dots.
The article follows a team from a community college in San Francisco that competes with the "big boys" -- I love the fact that one of the real powerhouses, from Slovenia, gets all kinds of local media attention.
Gene Expression's p-ter makes an interesting point about weak genome-wide associations and drug development.
Any doctor knows where I'm going with this: one of the best-selling groups of drugs in the world currently are statins, which inhibit the activity of (the gene product of) HMGCR. Of course, statins have already been invented, so this is something of a cherry-picked example, but my guess is that there are tens of additional examples like this waiting to be discovered in the wealth of genome-wide association study data. Figuring out which GWAS hits are promising drug targets will take time, effort, and a good deal of luck; in my opinion, this is the major lesson from Decode (which is not all that surprising a lesson)--drug development is really hard.
Yes, figuring out gene functional networks is the hard part; also, how alleles may interact in unexpected ways with different genetic backgrounds.
A pair of articles in my browser tabs refer to bioethics.
Ronald Bailey, in Reason, writes about the "ethics" of life extension research:
"How dare you do this research? The earth is already being raped by too many people, there is so much garbage, so much pollution."
Ten years ago, an anti-aging researcher described this hostile reaction to her work in the pages of The New York Times. Not much has changed since then.
I had exactly the same reaction from my undergraduate students last time I taught my anthropological genetics course. Sure, they said, people might like to live longer. But wouldn't that be a bad thing? The world is too crowded as it is.
I'm thinking that death was far beyond the horizon of their bright young minds.
Much of Bailey's article describes the way ethicists try to put a numerical value on happiness, multiplied by a number of years. It's not different in principle from an economist estimating the financial damage of an early and unexpected death. But somehow it seems laughable to me -- as if an individual's happiness were the only important variable. What about the value of grandparents to their descendants, or the value of living history to the whole population?
This happened to hit my desktop at the same time as Sally Satel's article, titled "The limits of bioethics".
She describes the history and scope of bioethics. Satel points out that the name "bioethics" was coined by Sargent Shriver, amid a burst of interest in the problems of biological and medical decisions in the late 1960's. She mentions the establishment of think tanks and expansion of the bioethicists' brief during the 1970's and 1980's, and touches on the controversies over "conservative" bioethics during the last decade.
What is the proper role of ethicists in decision-making? Here's Satel's conclusion:
At their best, bioethicists are scholars who study the intellectual and social history of value controversies in medicine and biotechnology. They can teach us about the technical and cultural antecedents of modern debates and show us how to engage in disciplined moral inquiry. They are skilled at drawing conceptual maps of the dilemma at hand while enumerating various ways to resolve it. In these ways, bioethicists have much to offer. But beyond this, their value is mainly cosmetic or bureaucratic. When called upon by politicians, their main task is to neutralize explosive issues or to provide ethical cover for decisions that have already been made. When physicians summon them, it is mostly to mediate disputes between patients, staff, and family members regarding end-of-life decisions. The media tap bioethicists for high-minded sound bites. In hospitals and in governmental agencies, they man the regulatory ramparts.
Maybe some bioethicists would disagree, but I think most see themselves as scholars instead of apparatchiks.
Science gives us a "policy forum" this week on gene doping. The lead author, Theodor Friedmann, is the chair of the "Gene Doping Expert Group" at the World Anti-Doping Agency.
The essay describes the current potential of gene doping, and is a good short review. I found the paragraphs on "marketing gene doping" very interesting:
Athletes are an especially vulnerable population in the marketing of performance enhancement (31). Reputable athletes or coaches with little knowledge of genetics are at a disadvantage in assessing "scientific" claims that appear in advertisements. Marketing is particularly worrisome when the science is still a work in progress, when a person's health can be adversely affected, and when consumer knowledge about genetics is low. Although advertisements promoting products that promise to enhance athletic performance have pervaded the Internet for many years, recently it has become home for advertisements that promote products to "alter muscle genes...by activating your genetic machinery" (32), or that state "your genetic limitations are a thing of the past!" (33) or "Finally, every bodybuilder can be genetically gifted!" (34).
These are people who make easy targets for nutrigenomics and other questionable areas of "health enhancement." A large number of amateur bodybuilders and fitness enthusiasts create a "gray market" supporting questionable products, and these products themselves support the ecosystem of effective performance enhancement drugs. They also create a lot of biochemical noise for those who want to find new tests for performance enhancers.
The essay includes a few paragraphs that describe the prospects of future detection of gene doping. It may be very difficult. Detecting some synthetic performance-enhancing agents requires the cooperation of primary producers, who add tracers to their products. Gene doping might enhance performance for periods months or years after the vector is administered, and may not require dosage that would significantly alter isotopic or chemical signatures, even if they contained such tracers. The authors suggest that the incidental biochemical effects of gene doping might enable the identification of a "signature" of such products:
For instance, exposure of murine myoblasts to IGF-1 has been shown to induce transcriptional and proteomic changes that may eventually constitute a "signature" specific for exogenous IGF-1 exposure (29, 30). Of course, the application of these kinds of global assays would require rigorous validation of a connection with specific doping agents or methods.
I am very skeptical -- it's likely that the "signature" of a doped individual will not differ appreciably from the normal variability of tissue metabolism, particularly in the subset of high-performance athletes. Still, it may be possible to find one particular tissue type that provides a high-information-content message about normal versus doped processes. I just think that a lot of innocent athletes are likely to get snared in this net as it closes in on clinical validation.
Friedmann T, Rabin O, Frankel MS. 2010. Gene doping and sport. Science 327:647-648. doi:10.1126/science.1177801
Data point:
This is the time in my introductory class when I discuss genetic disorders, and I described the new Counsyl test as part of my lecture today.
I took a quick poll -- how many of my undergraduates would be interested in getting a test that told them their carrier status for a hundred genetic disorders. More than half of them expressed interest. That just seems huge to me -- I don't think I'd have seen the same result ten or even five years ago. There's been a real change in attitude about genetic testing in the last few years as technology has gotten cheaper and has more public attention.