Hacker News Reader

Comment by Boogie_Man | Hacker News Reader

Can anyone vouch for the validity of nebula's privacy respecting testing? I've told people around me that I'd get it done as soon as I could trust they wouldn't sell me data. AFAIK I'm just German and Irish, but I've been mistaken as Turkish, Hispanic, and Chinese at different points in my life (by members of these ethnicities). I also have heterochromia, and worry that it might be related to inbreeding.

View on HN

If you want to sequence your genome entirely privately, the only real option is via https://store.nanoporetech.com/eu/minion.html (i.e buy a small sequencer and do it at home). Be warned that this will be expensive (more than the listed price here, e.g you also need a pipette etc.) and very much trial and error but ultimately it can be done at home.

I'm not certain my social reputation can handle asking friends and neighbors if they'll float me $150 to privately sequence their DNA in my basement but as long as I don't break out any skull calipers it might be fine.

No but I can attest to my friend using Dante labs. Nebula and Dante as well as all others outsource the lab tests and buy"bulk vouchers". The privacy isn't an issue, Dante has been known to hire really bad labs that lose or mislabel data and latness is a problem. If you choose one for privacy, they'd be the best since they are the worst.

Mislabeled data does not help you with a data set that can be fully attributed to you using >>1% of the entire data set.

The data isn't stored by the labs (not Dante, the testing lab) and, and these mishaps happening is further evidence it isn't. You don't need to have Dante keep your data after transfer and the information gleaned isn't that powerful or useful. Full genomes of people, a DNA dox is available from many people and you can download all several gigs of their data. Good luck doing something with it, it's like LLMs but more random, larger files, and hard to find definite information even from GWAS.

Turns out I am a lucky guy (that has found quite a lot in anonymous genome data set). I am a bioinformatician and have worked with population sequencing data set like gnomAD. There is a huge inherent value in genomes without metadata and we (the research group I am working in) are using them every day. There is a large world beyond GWAS.

Do you think that people should be concerned with their privacy?

I am definitely more on the privacy verse side, and do not personally know a single person working in genomics that got their genome sequenced. However, there are so many people now which did. So I do not think it can be a huge concern as too many would be affected if it was a big issue at some point (and technically, so one does really just need a little more that a hair from you to be able to get the same info). In addition to that, I think there is a lot more benefit in those data than concern.

However, this data is ultimately valuable. Insofar it makes sense to me that who ever has such data will have incentives to use it for their own benefit, whatever that may be. Its not so much 'I have nothing to hide', but rather 'I do not want you to use my picture without my consent'.

I spoke to razib Khan a bit and he was the only sample from bengal for a while, to be used. He was also the first that sequenced his kids DNA in the womb.

I knew someone who worked at Grail, the problem with government is they're too conservative with the data usage. Nothing as useful can be gleaned do people choose to dump their data into private databases to get caught up on research.

I don’t know how nebula is these days, but I had the worst time with them. I sent my kit I and six months later they had nothing and they kept telling me it was coming, It was coming until they said there was something wrong with it and I would have to do it again.

I asked for money back and they delayed and delayed and delayed until I said I was going to sue them when they finally gave me my money back.

Reply on news.ycombinator.com