I respectfully disagree with the claim of @manoliskellis that one should be comfortable putting one's genome on the web if one is comfortable searching the web. US statutory protections against genetic discrimination are woefully inadequate.

via @YouTube

GINA prohibits genetic discrimination in employment & health insurance, but has no prohibitions against genetic discrimination in vital areas like mortgage lending, education, disability insurance, or life insurance. Only a handful of states have more stringent protections.

Lest one think that this is all worrying about a low-probability event, well, disturbing cases of discrimination have already happened and have been litigated. In employment, see Norman-Bloodsaw v. LBNL and USEOC v. BNSF. In education, see Chadam v. Palo Alto Unified School Dist.

What about life or disability insurance? Well, it has happened there too, as well as discrimination based on familial history of a genetic disorder. See the cases of Jolene Hollar and Terri Seargeant. https://law.stanford.edu/wp-content/uploads/2018/03/slaughter.pdf

There is significant interest in state legislatures around the country, in bolstering genetic anti-discrimination protections. If the risk is less than that of "web searching", why would so many state legislatures be mulling strengthening the protections? Why do we have CalGINA?

The claim implicit in that video snippet is that there is very little predictive power in having one's genome online. But that is not true for certain rare variants with known clinical consequences, as well as rare variants with clinical consequences that are yet to be discovered

Whether having one's genome publicly available today has much predictive power is beside the point---companies will try to use it in risk models unless (and perhaps even if) it is illegal, whether or not it is scientifically justified based on current knowledge of human genetics.

The key point is that once the data brokers get their hands on your genome or your SNP genotypes, it cannot ever be made private again. And that risk to you of discrimination can only compound over time as predictive models improve, unless/until federal protections are bolstered.

I care deeply about genomics. I gladly use 1KG and dbVar data in research, and I am grateful for them. But I feel an obligation to be as clear as I can to the public about my assessment of the significant unknowns in the future risks from public genome data (identified or not).

For sure, I have deep privacy reservations about DTC SNP genotyping. Sure, they only batch-up and rent out their data, for now. But it's only a matter of time before one of these DTC genotyping companies goes Chapter 7 or is bought by a PE company and sells to a data broker.

Should you participate in a genomics project for the greater good? It's up to you. Read the consent form carefully. Consider what may change in the future (including the state you live in). Consider that you are providing genetic info about not just yourself, but your relatives.

And carefully weigh the possibility of a re-identification attack. Remember that even if your name is not on your genome data, a bad actor may be able to identify you by the similarity of the genome data to *identified* genome data from a cousin or sibling, in another database.