Concerns about 23andme testing

[Snow Leopard]

I really like the idea of SNP testing and the pricing of 23andme seems right, but I have serious concerns about their whole business model, particularly their lack of privacy protection and the fact that they (not you) apparently own the data that is generated.

I'm not really sure what to think about this overall and whether to go ahead?

http://www.theverge.com/2012/12/12/3759198/23andme-genetics-testing-50-million-data-mining
etc

 

[August59]

The whole SNP field seems like a lot of unverified data. Seems as if there are discrepancies between 23andMe, Genova and Yasko's meanings of certain SNP, alles or whatever some of them are called.

It just bothers me a little bit as I would hate for someone to interpret something wrong and send them into a tail spin.

They sure could take this data at a future time and write a book that would probably scare the hell out of a lot of people.

 

[Snow Leopard]

Oh, most of the so-called correlations of the SNP data are based on poor science or are not replicated, but that doesn't bother me so much as the science can catch up. The problem for me is the privacy and 'ownership' of the data. I believe that I should own the data, not them. If they want to use the data for research or whatever, then they should be licensing it from me, not the other way around.

 

[Valentijn]

I have serious concerns about their whole business model, particularly their lack of privacy protection and the fact that they (not you) apparently own the data that is generated.

When signing up (or at any other time) you can select to not have your results used in research. As the article linked above notes, you need to opt in to have your results available for research.

Yes, you are the product (if you're okay with that), but you can opt out and still get your full results. I agree that them never deleting their records of your genetic data is a bit dodgy, but they're not allowed to use that data if you don't let them.

And to some extent, I can see why we're not allowed to dictate what someone else erases from their computer. Looking at non-health comparisons - would it be reasonable to demand that our banks or creditors erase all data they have compiled about us? Certainly not, but the comparison is a bit weak, since health data is seen as being far more personal, and something which we played a big part in "producing".

The article has a very good point when it says "It’s hard to say exactly how those rights will play out, however, since none of this has ever gone before a judge." It's really undiscovered territory, with no clear parallels to be drawn from other areas. If 23andMe were doing the ethical thing, they'd delete their copies of patient data upon request - it would look better, and probably very very few people would request it, especially if the process were a pain in the butt.

But they are a business, and will ultimately do the most profitable thing, rather than the most moral thing. And failing to delete results is probably not eating into their profits to any significant amount. Hence if change is going to come, it will almost certainly be from the results of lawsuits and/or legislation. It's a bit distasteful, but it's exactly what every other major corporation does.

For myself, I found it to be highly worthwhile. If using my genetic data and questionnaires as a product to sell to companies means that we can continue to get great results for very cheap, I'm pretty happy to support that. I also think it's sufficient that people can opt out of having their data used if they want to, even if that option isn't ideal.

 

[Valentijn]

The whole SNP field seems like a lot of unverified data. Seems as if there are discrepancies between 23andMe, Genova and Yasko's meanings of certain SNP, alles or whatever some of them are called.

A lot of people really do not understand how to use their results, once they have them. The 23andMe interpretations are quite minimal, and generally of very limited usefulness.

What is needed is more programs which take the raw data generated by 23andMe, and apply it in various ways. One way is to pull out results known to be relevant to certain diseases or biological processes, such as is done by geneticgenie.org, based on Yasko interpretations.

A major problem arises when those interpretations are not based on credible science. But that is likely to be resolved as more reputable services are implemented, and users become aware of their options and the flaws with some of those services. Another problem exists in that the magnitude of the problem caused by the SNPs are either implied or inferred to be different than it is, but again this is easily fixed by citing to good research.

Rare allele analysis can also generate somewhat short lists for people who are sick, but who don't really know why they're sick. In the case of an ME/CFS patient, this might help identify things which are going wrong in regards to disease symptoms, or even identify possible alternative diagnoses. Then, in many cases, it is known how to address problems (such as megadoses of vitamin D for dealing with extreme VDR mutations, etc).

And on a related note, people with poorly understood yet fairly identifiable disease (such as ME, where underlying causes are a mystery, but some symptoms are pretty unique), patients with the same symptoms can compare their results to see if there are shared abnormalities. This can be done most easily with short lists of rare results, but would require a decent database setup and storage capabilities to handle a lot of full 23andMe files. But the ultimate outcome could be better understanding of the disease, and even a biomarker for having a predisposition toward developing the disease.

 

[August59]

Thanks guys for helping me get some more understanding of what is going on with this science. I agree with "Snow Leopard's" concerns about the personal data belonging to the individual. Even if someone "opted in" the data doesn't like it could ever be used in a credible study to try and replicate the results, unless when purchasing one of these test there was a comprehensive questionnaire of who this "sample" really is.

I think I will wait awhile to see how this science evolves. It's possible in the final scheme of things that the data that appears to have the most meaning now may not be very relevant if they ever find out what processes within our body are affected by this disease and the sequence it manipulates the body functions from the time it begins to the point in time that any one person is in the cycle. There are going to be patients with ME/CFS that at any point in time could all be at a different point in this disease and all will probably require different treatments, at least initially.

This is just my feelings on this disease as I have never in the 10 years of having this disease ever seen 2 people that have started and progressed the same way. I hate to sound negative, but treatments may end up being harder to find than the cause.