[mattwrench]

We may be capable of sequencing DNA cheaply enough for the average consumer, but what's the point? Issues like the missing heritability problem and an incomplete understanding of protein folding show we lack a full comprehension of the information DNA tells us. Cancer-killing bacteriophages are still years from being relevant as a treatment for the general public.

The Human Genome Project has yet to provide substantial medical advances. Genetic sequencing only seems useful to look for diseases with an explicit genetic link like Huntington's, and many people aren't going to want to know they have diseases like that. Otherwise genetic sequencing seems as useful as a full-body scan.

[ABrandt]

Although everything you say is true, there is a very clear benefit of genetic testing that exists today. Many drug interactions have a strong genetic link. Take the common blood thinning drug warfarin for example. It's considered a medical best practice to prescribe this drug to patient's with a variety of heart problems (irregular heartbeat, history of heart attack, etc). A single dosage has wildly different effects on different patients however. The same dosage can cause fatal bleeding in one patient, while in another patient have no effect at all. Without genetics, the proper dosage is a crap shoot. With genetics, it's an exact science (as it should be).

And this is just one single example. Research is revealing more genetic-drug links all the time with relatively small sample sizes. Genetic sequencing may not cure every disease like we were promised ~10 years ago, but it is saving lives and is essential to the advancement of medicine.

[jballanc]

This really is the key point. In graduate school I worked along side many individuals with close ties to the pharmaceutical industry, and many had very similar stories to tell...

Getting a drug approved is an interesting process. Not only do you have to prove your drug is not harmful, not only do you have to prove that it does what it's supposed to, you also have to prove that it works better than all the available alternatives. I heard numerous stories where a drug would go to trial, and for some subset of the trial population it would be a miracle cure. However, the way the current approval process runs, those miracles have to be averaged out with the rest of the population, and often the end result would be "it's not better than existing drugs: denied!"

Of course, scientists being the curious type quickly figured out that what they were seeing was the consequence of genetics. Have a certain variant of gene X? Then this is a miracle cure. But only 10% of the population has gene X, so on the whole this drug doesn't appear to be better than the alternatives.

The problem is, the FDA doesn't know how to approve drugs that only work for people with a specific variant of gene X. Mostly, it's a chicken and egg problem: most drug trials don't include genetic profiling, because it would be wasted money since the FDA doesn't consider genetic profiling, because most drug trials don't include genetic profiling.

If sequencing becomes cheap enough that it can be included as part of the standard drug trial process, this could all change. Personalized medicine is the future.

Edit: Interestingly, though the name escapes me at the moment, I did hear of one drug that was approved for a specific gene variant. However, this was only because that gene variant was particularly prevalent in African American men, and demographic data is collected during the trial process. I recall there were a lot of upset scientists regarding this outcome, though, because conflating race with genetics is dangerous and irresponsible. There might be an 80% correlation (and even that might be on the high side), but that implies that there are non-African American males who could benefit but won't get the drug, and African American males that will get the drug even though it has no beneficial effect...

Edit, post-Google: Well, that was rather easy to find -- http://www.nytimes.com/2005/06/24/health/24drugs.html

[jesseendahl]

BiDil actually is not a great example, because... "The trial, however, was conducted only in African American patients, and the results, therefore, give the impression that BiDil works only in African Americans. This is not the case. The trial investigators themselves concede that BiDil will work in people regardless of race." The drug company just pitched it that way to get it approved by the FDA.

Source: Kahn J (2005) From disparity to difference: How race-specific medicines may undermine policies to address inequalities in health care. South Calif Interdiscip Law J 15: 105–130.

http://128.125.42.47/why/students/orgs/ilj/assets/docs/15-1%...

[apaprocki]

Warfarin sensitivity is one of the more concrete useful things to come from being tested. I, for instance, have a genetic increased sensitivity to it and knowing that makes me feel better in case I ever need it. Apparently I also have substantially increased odds of liver toxicity from the osteoarthritis drug Lumiracoxib. Not immediately useful, but good to know. Currently 23andMe lists 20 drug responses based on the SNPs they test and existing research.

[Daniel_Newby]

> With genetics, it's an exact science (as it should be).

The therapeutic window for warfarin is strongly influenced by diet and gut microbes. Knowing patient genetics does not help.

[apaprocki]

According to this paper[1], prior work has already established that 42% of the dose variance is controlled by 3 SNPs. It isn't 100%, but 42% is not insignificant.

[1]: http://www.ncbi.nlm.nih.gov/sites/entrez?cmd=Search&db=p...

edit: Also on 23andMe's page regarding Wafarin:

  In January 2010 the FDA updated warfarin's label to say
  that information on these variants can assist physicians in 
  selecting a starting dose of the drug. The agency also 
  provided initial dosage recommendations for patients with 
  different variant combinations. The FDA does not, however, 
  require that genetic testing be done before prescribing
  warfarin.

[zekenie]

Agreed. Genetics is overrated and poorly understood. To be sure, nucleotide sequence is important in determining phenotype, but since the human genome project's findings were released, its become more and more clear that its no silver bullet. Protein expression, folding, and regulation are really complex systems. Very few pathologies actually come from DNA sequence. There's an entire emerging field, Epigenetics, which is basically heredity that is not the result of DNA sequence.

That blood thiner example brings up some interesting thoughts. Have they found links between specific sequences and the response to the drug? If so, this could be useful. If it is indeed really cheep to screen, it could be useful epidemiologically. It could find new associations between sequence and phenotype that weren't possible before. For example, if everyone's sequence was known, they might find that people with a certain copy of a gene didn't respond well to a certain class of drug. But, the fruits of this would come after everyone was sequenced, and after much study.

It also brings up some ethical concerns that you mentioned. What if you don't want to know you have Huntington's disease, or you are predisposed for heart disease? Or moreover, what if you don't want that information to be in a database somewhere for who-knows-who to access? These are complex issues.

[rickyconnolly]

The Human Genome Project wrapped up in 2002-2003. Since it takes about ten years to get a new drug from benchtop research to the doctor's office, we should be only starting to see the first trickle of drugs reaching the market now. The biopharmaceutical industry is mostly focussed on recombinant antibodies at the moment, but expect to see the first genetic and stem cell therapies gaining approval in the next 5-7 years.