Tuesday, August 11, 2009

Will you buy your genome on a disk?

A story in the New York Times today describes a new DNA sequencing technology that will sequence a whole genome for under $50,000. This is lot closer to the $1000 genome that researchers have been waiting for (and promising) for a long time, as a research tool, but more importantly as an invaluable tool in diagnosis and prediction of disease risk. The grand 1-grand for 1-genotype idea is sure to become a reality sometime soon.

But, even science writers who are outliers on the genetics hyperbole scale are now routinely aware and questioning what we can gain from this information. Essentially, whole genome sequences extend association studies because they are looking for variants in sequence that correspond statistically to variance in phenotypes like disease. The scaled-up GWAS and big biobanks will be the sample on which such work will be done.

As regular readers of this blog know, we (and we're not alone) have been questioning the meaning of 'genes for' thinking for a long time--is this now percolating into the general consciousness even in the media?

Well, there are huge vested interests hiding under the bed. In spite of what looks to be an increasing acceptance of genetic complexity, adherents of 'genes for' thinking are still spending increasing time and money on genome-wide association studies (GWAS), looking for genes for their trait, and still claiming great success, DNA testing companies like deCODEme and 23andMe are still in business, claiming to be able to tell you your risk of disease, and people are still buying these services.

Those who are not so savvy but need to keep their careers on track, and who can do these kinds of studies (because they are largely canned and off-the-shelf nowadays), are sometimes perforce committed to this status quo. But for various reasons that range from true belief in the prospects to fully aware budget-protection are pressing ahead. They need the funding to continue to flow, and hope or believe that whole genome sequences will save the day. Somehow. They don't know how. Pray for serendipity!

It is easy to criticize and harder to change course, especially with so much invested in equipment, equipment manufacture, bureaucratic portfolios, lab personnel, publications, reputations, and tenure. In this sense, we think science is forced to stay the course since we're only human. But that doesn't make it the best science, even if it's technologically leading edge and extremely sophisticated, which it is.

To be a bit more sympathetic, most people are rather conventional and conceptually not very innovative. In science as well as other areas of human endeavor, we want our ideas and even our dogmas: they give continuity to our lives and a sense that we understand things. Change comes hard and new ideas even harder. Though we're all taught, and many teach, that the objective of a scientist to prove his/her ideas are wrong, that's near-total baloney! What is done is almost always contorting to prove that our ideas are right. That's how careers are built. Many journals won't even publish negative results. That's why even in the face of negative results, as in this case, we persist. But that doesn't make it good science.

As to the promises that genomes will predict your life experience....we're not buying it.

4 comments:

RBH said...

"In spite of what looks to be an increasing acceptance of genetic complexity, adherents of 'genes for' thinking are still spending increasing time and money on genome-wide association studies (GWAS), looking for genes for their trait, and still claiming great success, DNA testing companies like deCODEme and 23andMe are still in business, claiming to be able to tell you your risk of disease, and people are still buying these services."

And homeopaths and tarot readers are still in business. :)

Ken Weiss said...

But they don't have billions of public dollars, and venture capital, being invested in them. How different are they from, say, Freudian psychology, phrenology, or alchemy in the sense that some of these things that have weak support become established as legitimate, but others don't?

It's not unusual. Basically unsupported theories were accepted doctrine in the physical sciences, and humoural medicine lasted 2000 years!

In all cases,and probably even in homeopathy and tarot, after the fact there will always be some apparent successes and that keeps them going.

In genetics, there are also successes that are touted after the fact, a kind of conditional post-hoc evaluation that can be used to support almost any view, no? It only takes an occasional such 'hit' to reinforce a belief one doesn't want to give up.

zejen said...

So, what makes good science?

Anne Buchanan said...

Zejen, with respect to the issues we wrote about in this post, good science involves understanding what genetics has taught us over many years about complexity. The promises that many prominent geneticists have made over the years, that knowing our DNA sequences will let us prevent whatever diseases we are genetically programmed to have, and we'll thus stay healthy into a ripe old age just aren't true.

This isn't a secret. That most traits and many chronic diseases are due to many genes, and/or gene(s) by environment interaction, is well-known, but still too often ignored. And, even if we do know what environmental triggers lead to disease, environments can't be predicted, and so neither can future gene by environment interactions.

So, our point is that this knowledge is out there, but that too many geneticists continue to ignore biological complexity in favor of pursuing 'genes for' their favorite trait, when paper after paper after paper shows that this is not a successful strategy.