The Awl–which I totally and unabashedly love–has a piece on the over-promise and under-delivery of personal genomics. It’s a familiar meme and probably not news to Genomeboy readers (both of you!) that a fair number of folks feel this way. I have some sympathy for the “too much hype” argument, especially in light of how complicated common diseases and traits have turned out to be and how the personal genomics industry (if we can still call it that) has declined to come together to develop standards, choosing instead to fend off the annual sabre-rattling by the FDA on an ad hoc basis.
As I recall, I spent a long time on the phone with the Awl writer and said many things, including this:
“It turns out to be just a total fucking mess. So instead of having this linear icon representing human biology, the most potent symbol now is the hairball.”
If we are talking only about type 2 diabetes or human height or Crohn’s disease or certain forms of cancer, then yes, I wholeheartedly stand behind the expletive and the hairball.
Personal genomics, in my view, should not be judged solely on its inability to deliver meaningful risk information about common diseases or how tall someone is likely to be. Our genomes harbor other stuff; we can choose to ignore it if we want to, but that doesn’t mean it ain’t there.
The two abiding criticisms of personal genomics are that the information is 1) useless and 2) dangerous. So here’s half a touché: I will go ahead and concede that knowing my lifetime genetic risk of developing primary biliary cirrhosis is 1 in 2000 (or 1 in 500) instead of the population average of 1 in 1250 has no substantive meaning for my life or how I choose to live it. Hell, it’s not even clear the extent to which genes contribute to PBC.
Which leaves the genome as “DANGER,” Will Robinson, something that was not addressed in The Awl article. My 23andMe profile includes information about my carrier status for 44 diseases caused by defective single genes. If those diseases are inherited in an autosomal recessive fashion, and if I carry one bad copy and my mate carries one bad copy, then our children have a 25% risk of being affected with that disease. And it’s serious shit like Tay-Sachs and cystic fibrosis and muscular dystrophy.
If we’re talking about hereditary breast cancer, an autosomal dominant condition, then a single bad copy of the BRCA1 or BRCA2 genes would mean something like an 80% lifetime risk of developing breast and/or ovarian cancer in female carriers. 23andMe tests for the three most common hereditary breast cancer mutations in Ashkenazi Jews. As it happens, I am of Ashkenazi descent and my Mom had early-onset breast cancer (she is alive and well and is wondering why you never call). Thus, when I learned from 23andMe and DNA Direct that I did not carry any of those mutations and therefore could not transmit them to my daughters, I was relieved. Fucking right I was! Say what you will about consumer genomics, but hereditary breast cancer risk information is not useless.
“But Misha, you’re a geneticist! You understand these things! We can’t disclose this sort of information to The Ignorant and Unsuspecting Public! They will jump off a bridge! They can’t handle this without a man in a white coat patting them on the knee!”
This is an old and tired argument and having to combat it makes me feel old and tired.
- No one is denying anybody the right to consult a geneticist. Be my guest. But understand this: if you want to see one in my town, you’re gonna wait six to eight months for an appointment.
- Consumer genomics customers are a self-selected bunch save for those people who get spit kits as Christmas gifts. What business is it of yours or mine if they want to know what’s inside their own cells in the privacy of their own homes?
- For the umpteenth time, read Robert Green’s work about disclosure of genetic risk for Alzheimer’s disease to asymptomatic people. We have no effective treatment for AD yet people still want to know their risk. And almost without exception they don’t regret the decision to know. They’d even pay for it.
- Last I heard 23andMe charges $99 plus $9/month. They sell information. That information is subject to change. When it changes they send you an email to let you know. Does your doctor do that? I kinda doubt it. But if you’re not comfortable with that type of uncertainty then you know what: don’t buy it. If you’re really ill and you want/need to see a geneticist, then do it. If it’s an unwanted Christmas gift, send your spit kit back and get something useful.
- Consumer genetics companies return information about drug response that, in some cases, is strongly influenced by genes. For example, if you have a certain genetic predisposition and take abacavir to treat your HIV infection, you could die. Not useless.
- Some of my postmodernist friends tend to look down their noses at genetic ancestry testing. I would argue that they are genetic determinists. Why assume that genetic information is so omnipotent as to irrevocably unravel one’s identity? Why must one narrative trump another? “Because it’s TECHNOLOGY! It’s GENETICS! It is ALL POWERFUL!” Please. It’s just another way of looking at one’s ancestry. And learning about genetics: I would argue that Henry Louis Gates has done as much to stir public interest in genetics as anyone or anything since the Human Genome Project. For realz.
So yeah, it’s a mess and a hairball and a bushel and a peck. And so not destiny. But that doesn’t mean that some of it isn’t useful and fascinating. Of course it’s fascinating: it’s about us.