3 ms·

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

by patall 3y ago

graphe 3y ago
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.
- patall 3y ago
  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.
  - graphe 3y ago
    Do you think that people should be concerned with their privacy?
    - patall 3y ago
      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'.
      - graphe 3y ago
        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.

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

by patall 3y ago

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

graphe 3y ago

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.

patall 3y ago
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.
- graphe 3y ago
  Do you think that people should be concerned with their privacy?
  - patall 3y ago
    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'.
    - graphe 3y ago
      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.