I don’t trust ancestry.com and don’t want to give them my DNA or anyone else for that matter.
23&me is obviously out of the question, so I’m wondering:
*Are privacy enthusiasts just locked out of knowing ethnic roots for certain?*
Is there any kind of technology that allows for a DNA background test without risking privacy? Perhaps some kind of [Homomorphic encryption]( Homomorphic encryption - Wikipedia ) that would allow analysis to be performed on “encrypted DNA"?
Sounds like far off impossible future tech, I know, but I’m just wondering what research is out there.
Has anyone put research into this?
I’d like to use one if these services, but I’m not willing to compromise my DNA information.
For ancestry? Not really, since the way this works is by comparing to other people’s results on a massive scale so… clearly data needs to be shared and accessed widely.
From a technical perspective I see no reason homorphic encryption couldn’t be used. But it is not used and I have seen no research into using it from anybody.
For health or other purposes that don’t have to do with comparing to other people’s data you could theoretically do this privately already, because there are standard formats for DNA/genome results that can be input into open source tools and run locally, and there are collaborative investigations like SNPedia which can tell you how certain results can be interpreted.
I have actually done this and have the only copy of that data on my NAS (this is like a 50GB file btw lol), but that was a unique situation. I don’t know of any direct-to-consumer ways to be tested and get this data from a trustworthy place (that won’t turn around and sell it) unfortunately, so it’s out of reach to most.
WOW, this is opening a rabbit hole for me! Would love to know how you acquired that data about yourself and how you’re certain that you have the only copy. That certainly does sound like a unique situation. Understandable if you’d rather not share.
Indeed, this is what I assumed. It would be interesting to see this market gap filled.
An open source collaborative ancestry wiki that uses homomorphic encryption to ensure user privacy.
I really don’t think it can be replicated by basically anyone unfortunately, so it’s not worth going into details. To be honest this data is also kind of useless information overall, because interpreting it is an entire challenge unto itself. Unless you are just a genetics nerd (like me) but this is probably outside the scope of this forum
I basically audited the whole process in person.
Yeah but hard to do without a medical prescription or a research/commercial affiliation. Even if you manage to get a prescription it’s generally unlikely you’ll be able to obtain access to the raw data. And/or it will be very non-comprehensive.
The biggest problem with this “product” is that I don’t think homomorphic encryption could be used to protect everybody’s data.
What it could be used for is to privately check your data against a public/unencrypted set of data. That data would have to come from somewhere, and the big players like Ancestry or 23andMe aren’t sharing theirs.
So the service would be stuck checking against probably inferior reference/academic data that becomes more outdated every day. And they can’t build their own database without being able to access their customer’s data. So… the general public probably just won’t be interested, and that’s why there’s no research into this subject. From a researcher’s perspective the only useful DNA data is data from a massive population, not one guy, so everyone is incentivized to not do what you’re asking as much as possible.
Overall genetic testing for consumers is pretty overrated anyways and more of a novelty than a real use for most people, so it’s easiest to just recommend not doing it at all.
As a recreational activity, with no serious consequences at stake, it barely matters whether the results consumers receive from their DNA ancestry tests accurately represent the percentages of their ancestry from different geographical regions. Given that there are no international standards for such testing, unlike genetic disease tests, it is not surprising that the results from different ancestry testing companies vary. As noted in Chapter 4, there are several stages in the analysis of a person’s saliva or cheek swabs where the criteria, reference frames, or analytics can vary among companies, yielding different outcomes.
I’d say a family tree from birth registries could give better insight but it only goes so far.
