> This article fails to mention any privacy risks to the volunteers.

What privacy risk, exactly? Who has been harmed? 23andMe has been in operation since ~2007 and between them and Ancestry.com and the UKBB and the Japanese BB and others, there must be somewhere on the order of 10 million+ genomes cumulatively. UKBB alone has been the basis for what must be at least hundreds of papers at this point by thousands of researchers or groups. Even further: there are literally thousands of genomes available publicly online from groups like PGP for a decade now, which you can go and download right now, including mine. Is all this not enough? When should we expect to finally see these risks materialize?

> GWAS have proven to be, at best, of little value. In fact, due to the unfortunate way that many researchers chase p-values, these analyses often result in misdirection.

That is completely wrong. GWASes have an excellent replication record, and were founded on the basis of stringent p-values precisely because of how earlier candidate-gene studies were garbage due to researchers chasing p-values.

> What privacy risk, exactly?

People will eventually be identified individually, as inevitably as the people in the AOL search leak years ago. Then they -- and their blood relatives -- will be associated with negative health traits.

And then, companies will discriminate -- in hiring, in insurance... anywhere they can. Because it will be financially advantageous to individual companies even though it is detrimental to society as a whole. Because companies are amoral organisms and it is folly to expect them to behave otherwise.

We're probably no more than a year or two away from a sequenced genome being considered PHI for exactly this reason.

It's always important to put this in perspective with the human History.

There is no place on earth that lived in peace for ever. Not even for a very long time actually.

Any country goes back to times were it's at war on it's own territory, or were it's under a dictatorship.

It will be true again for France, the US, Canada, Spain, any country.

We will go back there, it's an absolute certainty.

Maybe not (hopefully not), in our lifetime.

But ignoring this is not only wearing blinders, it's a recipe for making it worse.

Things like AI, facebook, the NSA intercepting anything, image recognition, cameras in every streets, tracking devices in every pockets and... DNA databases are going to be weaponized against the population.

Even if you don't believe in this (which is always strange to me, but for the sake of the argument, I'll have to acceptit), at least consider that it will make any neutral entity in power more likely to become malevolent by giving it the opportunity to access abusable huge power.

Hence, if we want to survive and evolve as a specie, we need to deal with those tools carefully, as they are very potent, and this means what they can do for us is as great as what it can do against us.

This is a super interesting discussion. Both sides make good points. I wonder if in the future cost and technology of sequencing will evolve so much that somebody will start harvesting as much data as possible from places like metro, train stations, malls and so on. This would make it inevitable that your DNA ends up in these databases and you just have to accept it. Just like currently one kind of has to accept that thanks to facial recognition getting even better and cameras everywhere one can be tracked even without any electronic devices.

You don't have to accept it. You can protest. You can explain. You can vote. You can paie for service respecting you. You can educate. Maybe there are others things to do.

There is always a choice. Most people saying there is none are just choosing one cost against another and say it's not a choice. Not blaming, i do the same.

But we could avoid the disasters, living by thinking about the society ou simple everyday actions produce.

However it's not only very hard to do, but most people are not interested in doing so. Or even to consider the question of society, or how to live ones live.

Living on auto pilot is one of humanity's biggest problem. It helped us a lot when we were in survival mode, but now that our own construct is the only real danger to our existence, it really is an hindrance.

If something is possible and you can't prevent it from happening then accepting it seems like the best option. You're arguing that you can prevent it from happening by regulation and education. I don't think it works if you apply a security mindset. Mallories don't care about rules. 3 letter agencies would be just one of examples.

Yes, you can lower the probability of exploitation happening, but you have to keep in mind that it is possible. Kind of like leaving your car open - it's quite unlikely that somebody will go and pull the handle checking if it's closed, but it's possible and you're probably better off taking that into account.

(now that I realized you are also parent of my original reply I'm a bit confused, I mean your two replies are kinda responding to each other)

See the history of credit reporting in the US for a counterpoint (Consumer Financial Protection Bureau, Fair Credit Reporting Act, etc.).

I expect that companies won't be able to help themselves and that they will eventually spasm in abuse of ever-more-accessible personal data so egregious and evil that people will take notice.

If you can’t prevent it is a higher standard than most make it out to be. For most it’s “can’t prevent it without discomforting myself,” which is no standard at all.

These data can be used to identify their source. This will become easier as more data becomes available.

https://news.ycombinator.com/item?id=15209306

I guess the value of the data depends on who measures its worth. Replicated p-values lead to multiple publications, which are valuable to academics. But these results can still cause misdirection. What is a low p-value worth if it doesn't lead anywhere?

> These data can be used to identify their source. This will become easier as more data becomes available.

Yes, so? (The face paper you link isn't a good example of this anyway, since it's so questionable and re-identification already works much better than that based on stuff like haplotypes.) That's not a real risk, not compared to, you know unlocking the secrets of human genetics. And it has little to do with the research databases: DNA is inherently as far away as it is possible to get from being anonymous, once it started becoming cheap, the loss of genetic privacy was inevitable no matter how it was handled. And it's not that big a deal. We will be much better off for it.

> Replicated p-values lead to multiple publications, which are valuable to academics.

This is very different from what you were just claiming.

> That's not a real risk, not compared to, you know unlocking the secrets of human genetics.

So, it's you who gets to play god and decide to sacrifice individuals for the alleged greater good?

Unfortunately, GWAS data is unlikely to unlock the secrets of human genetics. We have been in the "post-GWAS era" of genetics for at least a few years now.

> And it's not that big a deal. We will be much better off for it.

I hope you are right! However, I have doubts. Also, just because sensitive information can be compromised doesn't mean that it should be offered freely.

Using the information to assist development of targeted pathogens for biological warfare is an extreme example.

> GWAS have proven to be, at best, of little value. In fact, due to the unfortunate way that many researchers chase p-values, these analyses often result in misdirection.

I would certainly challenge that, its just that some people considered GWAS as the solution for everything which it clearly isn't, especially because rare mutations where underestimated (and cannot be discriminated in GWAS) and because GWAS was never able to detect the causing variant, just associated/linked genomic loci. And last but not least, the regulatory genome is turning out more and more complex (with interacting variants leading to complexities of O(x^n)) and we do not know yet how it will all be understood, its just that GWAS is one of the better tools we have.

Not be the case for everybody, but I'm 100% fine with having my genomic data public, especially in the context of research. That's not to say there aren't potential societal contexts where it would be totally awful (e.g. 1930s Germany) but I can't imagine any risk to myself.

But thats not the point. The question is whether you can make data public that, while it belongs to you, also belongs (partially) to other people (your siblings, parents, unborn children ...) who might not want to have their genetic data public. Sure, there is may be no risk to you and a data repository like the one described contains an enourmous value for society in the form of biomedical data but that does not mean you have to give it to everyone without knowing their intention. Especially if they end up knowing something about you that you don't.

This is an incredibly important point!

Openly sharing your genetic information makes both you AND your relatives more vulnerable, with vulnerability decreasing as relatedness decreases. It also reveals potentially sensitive information about family traits.

That's one (rather selfish) way of thinking about it. Another way of thinking about it is that you (and everybody else) is better off if more genomic data is available for research. It's worth remembering that genomicists like George Church and Craig Venter have made their genomes publically available for all to access.

I think i have the right to do anything i want with my genome, including making it public. It's my genome ffs, it's me. If i don't have this right with my genome then I don't have any rights at all.

Except it really isn't entirely your own genome. You share most of it with your relatives, and to lesser extents, the rest of us and all other beings. Since it's a shared resource, they should have a say.

Can you follow the argument if we flip it around? Your right to do anything you want should include not making it public. Others publishing "theirs" would mean publishing yours, whether you want it or not.

It is entirely my genome.

Over 99.9% of the genome is identical to every human being. And that genome is made public (the reference genome) and updated regularly as we understand more about the DNA.

Even identical twins don't have the same DNA. So it's not like I make public another person's DNA.

Sorry but no, just because something is yours it does not mean you can do with it whatever you want (I understand that this is a totally non-libertarian view). If you find uranium on your plot of land, you also cannot sell it to whoever you want. And the human genome is similarly seen as something protected, just look at how CRISPR experiments on human embryos are highly controversial.

> And the human genome is similarly seen as something protected

It's not, I can publish/sell/do whatever I want with any of my genomic data I have desire to

Of course - this is total bullshit - trying to tell people what they can do with their own dna.

Well, I feel sorry for your descendants who won't get hired when it's discovered that they might have inherited BRCA1 and Huntington's Disease from you.

Say you are a high risk for cancer insurance will be more expensive employers in the UA will not want to employ you.