[carbocation]

I'm disentangling these two issues intentionally. While the companies that have been offering paid genotyping services to the public have also been making, e.g., risk prediction from this data, it doesn't have to be this way.

I'm suggesting where I think the regulatory hurdle should be placed.

[prat]

sure - its the predicting thats causing the alarm, check this out.. http://www.gao.gov/products/GAO-10-847T

[carbocation]

My lab is a human cardiovascular genetics lab, so I'm familiar with these issues and I am also concerned. I read that report when it came out and was not in the least bit surprised.

I stand by my suggestion to disentangle concerns with giving someone their sequence (no concern--it's THEIR sequence!) from the concerns about companies offering quasi-medical services based on that sequence. The latter issue is perhaps analogous to practicing medicine without a license, although that is admittedly a bit of an overstatement since they're generally making predictions, not diagnoses. At any rate, the companies' behaviors are concerning.

[nitrogen]

With no disrespect intended to the doctors and researchers who comprise the medical field, I find it interesting that those closest to the field show the most concern over interpretation of sequencing results. As an educated medical layperson, I still believe that companies like 23andMe should be able to offer statistical interpretation of sequencing data. At the most, I could accept a required disclaimer, but it is a statistical fact that certain studies have associated certain SNPs with certain conditions, and I should have as much of a right to know those facts as I do to read the raw sequence of my DNA.

[carbocation]

> With no disrespect intended to the doctors and researchers who comprise the medical field, I find it interesting that those closest to the field show the most concern over interpretation of sequencing results.

That's not disrespectful at all. In fact, it's to be expected that those who are closest to a technique are the most familiar with its pitfalls.

> it is a statistical fact that certain studies have associated certain SNPs with certain conditions

Sort of, but that's not as relevant as you seem to think. The SNP reported in each study is not generally presumed to be the causal SNP; it's just tagging the causal SNP. For any particular person, it is possible that recombination occurred between the tagged and causal SNPs, interrupting the haplotype.

Even if we assume that sort of thing doesn't happen (which, most of the time, it won't), we're still talking about odds ratios of at most (for lipids) 1.3. That type of minimal increase in risk extremely difficult to convey. In fact, in medical practice, it often wouldn't even be worth reporting (depending on, e.g., background prevalence of disease).

So really, I think that issues are more complex and subtle than one might gather from reading your post. Should you have the right to get your genome sequence? Yes. Should you have the right to read the published literature? Yes. Does a company have the right to interpret that literature as a quasi-medical service? Again, this is the place where regulation makes sense.

[yummyfajitas]

If you want to shut down shady websites in China selling Lance Armstrong branded DNA tests, be my guest.

But what does that have to do with the regulations being proposed, namely requiring a doctor's permission/interpretation to get DNA tests from reputable sites like 23&Me?