3 ms·
I'd love to check my DNA for potential diseases, and I'd like to know, for instance, if I'm a descendant of Genghis Khan - but this article reinforces my reluct
by pavel_lishin 11y ago
I'd love to check my DNA for potential diseases, and I'd like to know, for instance, if I'm a descendant of Genghis Khan - but this article reinforces my reluctance to send it in anywhere.
As I understand it, they refuse to do anonymous testing - they have good reasons, some of which are completely focused on privacy. I'd love to have an at-home kit, so I could check the contents of my DNA against a database of known facts.
- curun1r 11y agoThere's always https://genographic.nationalgeographic.com https://genographic.nationalgeographic.com They won't give you information about potential diseases, but they'll give you information about your ancestry and ethnic makeup and they'll do it anonymously. They give you the option to register, but you can also lookup your results by the tag on your sample.
- jacquesm 11y agoDNA and anonymously do not go together.
- maxerickson 11y agoWhat use is the DNA match to third parties if there is no name/identity information paired with it? I guess it reveals a few bits of information if you know your killer submitted a sample to the project (but an easy way to avoid this threat is to not be a killer). Most of the threats I can think of start with the third party already having the DNA, so it's existence in the project database doesn't seem important.
- jacquesm 11y agoSomeone else that is a (distant) family member of yours could supply enough information that by combining a few of those records you'd be as identifiable as if you had added your name and SSN to the sample. It's not a problem in isolation, it is by combining data from several sources that you can re-establish identity from an otherwise anonymous submission. For instance, let's say my brother submits his DNA non-anonymously to some database. By looking at the match between his profile and mine it could be established that I'm a close relative. A few more like that and some constraint trickery and you'd have me as the only solution to your equation. For an interesting perspective on this: consider what could happen if Facebook bought 23andme... (not much chance, fortunately given who founded 23andme but still).
- maxerickson 11y agoThe relative match can happen whether you have submitted your DNA or not. The third party that is searching using DNA gets "Bill's close relative" or "Bill's close relative that also anonymously submitted their DNA to the database". The specter of the service working internally to de-anonymize submissions is real enough though. I guess I don't really care about protecting myself from future dystopias, not really in a sense that I have nothing to hide (I find attention quite uncomfortable), more in a sense that I hope I'd stand up and shout loudly, rendering irrelevant all the past hiding-planning.
- jacquesm 11y agoThere's approximately 100,000 Amsterdam Jews that would stand with you and should just as loudly, only they can't. Lessons from the past can only be applied to the future if you're actually willing to learn them. Unfortunately, it seems as though those lessons weren't serious enough yet. I shudder to think of the kind of lesson that it would take to get people to understand these things in a way that we will actually live by those lessons in the future. Standing and shouting loudly is not much of a defense against an organized entity that is 'out to get you', even if that seems like a distant and remote possibility, the damage it could do is sufficient to warrant the - small - premium we'd pay in being careful with our data today. It's a bit like the rest of the security industry: probability of incidence * potential damage is a good way to figure out whether or not something is worth protecting against. I wrote a bit more about this here: http://jacquesmattheij.com/if-you-have-nothing-to-hide http://jacquesmattheij.com/if-you-have-nothing-to-hide Because I got very tired of the 'if you have nothing to hide' mantra, and even if you freely admit to finding attention quite uncomfortable you might be persuaded by the argument that seemingly innocent entries in databases have cost many lives already.
- maxerickson 11y agoIs it okay if I think the better lesson (vs hiding) is to not allow tyranny? Maybe it's overly optimistic or foolish or something, but it's still a shit world if only the people that (properly!) planned ahead survive. Edit: Imagine a modern populist horror that subjected everyone to mandatory genetic testing and slaughtered people with certain genes; You can't escape your genes, the problem would be the populist horror, not whether you were on some preexisting list of people that needed slaughterin'.
- curun1r 11y agoThat may be true, but there are degrees of anonymity. The problem described in the article is not the DNA, but the association of the DNA with some other identifier. If all they have is a DNA sample, the only current way to look it up would be to match it against another DNA sample, which only tells you that you submitted a DNA sample. In the future, it may be possible to take a sequenced DNA and simulate what the person looks like, but we've barely crossed the threshold of having enough computing power to sequence a genome, so we're many years from that being possible and, even then, without knowing the donors age, you wouldn't be able to say what the donor looks like today.
- currymj 11y agoYou can still get a lot out of DNA today. There have already been murderers apprehended because their family members agreed to undergo DNA testing even though they didn't. IIRC it's possible in many cases to go from anonymous genome (and no other info) to a surname with a distressingly high degree of accuracy. It doesn't really involve trying to predict phenotypes or anything, it's basically just because you are distantly related with at least some people who have their names attached to DNA sequences. It becomes even easier to identify someone if you have some very limited extra information. Like, say, the postcode their sample was mailed from. My understanding is that basically the idea of an "anonymous" DNA sequence is going to quickly become meaningless, that unless the person being sampled takes extraordinary effort it will be possible to narrow down who they are. It only takes like 30-40 bits of information to uniquely identify a human.
- dekhn 11y agowe've had enough computing power to sequence a genome for 15 years now.
- 13of40 11y agoI have an uncle who got tested by a program like that a few years back. I'm not sure if it was the same one, but the results basically just said he was a male of Western European descent.
- Paul-ish 11y ago> I'd love to have an at-home kit, so I could check the contents of my DNA against a database of known facts. I think sequencing at home would be difficult. What we would need is for them to sequence your DNA, then send that data to you, then destroy their copy of the data. This would need to be enforced contractually. The problem is that if you lose your copy of the data, it is gone. This has issues with usability. If if they offered to keep your named detached from the sequence it wouldn't do much. Your DNA probably has more bits of entropy than your first and last name together.
- abecedarius 11y agoSequencing at home is done all the time by your own cells -- or anyway replication is. There's no inherent reason it can't be done by small cheap tools; we're just not there yet. (Yes, privacy is why I never signed up with 23andme. I expected news like this.)
- abecedarius 11y agoAdded: https://en.wikipedia.org/wiki/Nanopore_sequencing https://en.wikipedia.org/wiki/Nanopore_sequencing offers compact sequencers. Here's a handheld one: http://www.wired.com/2012/03/oxford-nanopore-sequencing-usb/ http://www.wired.com/2012/03/oxford-nanopore-sequencing-usb/ There can't be any physical barrier to high-quality readout by machines as small as a cell, because our own cells do it. I only brought that up because, ironically, I know more about cells than nanopore sequencing.
- damurdock 11y agoWe're still years off from sequencing at home for anyone other than dedicated biohackers. The MinION is really cool, but it's very much an advanced researcher's tool. Honestly, if you're interested in having your genome sequenced/analyzed, I would suggest contacting a local genetic counselor and asking if they can find a sequencing core (or someone that does beadchips) that would agree to delete your data after analysis. They would also be able to discuss your data and the impact it has on your health with you, and provide referrals if more testing or diagnosis is needed.