3 ms·

This is unnecessarily paranoid. It would take a concerted effort to obtain a good quality and sufficiently pure DNA sample from a person's ambient DNA. In the m

by Gatsky 8y ago

This is unnecessarily paranoid. It would take a concerted effort to obtain a good quality and sufficiently pure DNA sample from a person's ambient DNA. In the mythical scenario that someone wanted to get your whole genome sequence without you realising, they would need to collect many ambient samples (the gum you just threw in the bin, swabs of handles and keypads that you just touched, hair samples) and perform whole genome sequencing, which would cost multiple thousands of dollars. 23andMe's crappy SNP test is useless on a sample that could be contaminated with other human DNA.

rootusrootus 8y ago
Full sequencing is under a thousand bucks now and has even been on sale once for $200. Another year or so and I bet you will be able to have it done for less than $100
- Gatsky 8y ago
  My point is you would need to sequence multiple samples for the same person, and with contaminated DNA you need higher read depth which costs more. If you want to say that the person who spat out the piece of gum you sequenced is female and European, yeah you can probably do that for $200.
- btilly 8y ago
  Do you have a source? My understanding is that what 23andme et al do is very far from actual full sequencing.
  - dekhn 8y ago
    you can get 30X at $800 from at least one provider but that's not saying much, I think most people would prefer higher depth for accurate variant calling.
    - btilly 8y ago
      I am following this with interest. I have some variant of Marfan's syndrome but have not shown the associated heart problems. After consultation with a geneticist we concluded that I could spend a bunch of money to get the exact variant that I have of the relevant genes sequenced. But almost certainly I would learn that I have some variant of Marfan's syndrome that is not understood. Relatives who also have the same phenotype can get sequenced to verify that it is genetic. But since none of us are showing the most troublesome problems, and the literature almost certainly doesn't have our exact variation, I would learn nothing more than I know already. So just get an MRI every 5 years, and wait for costs to come down and research to improve. In 10-15 years look again and maybe the genetic test will be useful then. For a genetic condition as specific as mine, the sequencing done by these commercial providers is useless.
      - dekhn 8y ago
        I agree. I know a lot of genomics and health and after having my whole genome done, the results were worthless- I knew more than the genetic counsellors and had to explain to them that just randomly googling for SNPs and grabbinbg keywords from the research papers was a terrible idea. There are a few very well done genetic tests for a limited number of diseases.
        Gatsky 8y ago
        Finally, a voice of reason.
  - rootusrootus 8y ago
    Veritas does a full sequence for $999 and had a sale late last year for $200. I think Dante Labs will do it for $599. This is the whole thing, raw data included. Not 23andme crap. Edit: Dante seems to be doing the basic 30x full sequence for $229 at the moment.

by Gatsky 8y ago

rootusrootus 8y ago

Full sequencing is under a thousand bucks now and has even been on sale once for $200. Another year or so and I bet you will be able to have it done for less than $100

Gatsky 8y ago
My point is you would need to sequence multiple samples for the same person, and with contaminated DNA you need higher read depth which costs more. If you want to say that the person who spat out the piece of gum you sequenced is female and European, yeah you can probably do that for $200.
btilly 8y ago
Do you have a source? My understanding is that what 23andme et al do is very far from actual full sequencing.
- dekhn 8y ago
  you can get 30X at $800 from at least one provider but that's not saying much, I think most people would prefer higher depth for accurate variant calling.
  - btilly 8y ago
    I am following this with interest. I have some variant of Marfan's syndrome but have not shown the associated heart problems. After consultation with a geneticist we concluded that I could spend a bunch of money to get the exact variant that I have of the relevant genes sequenced. But almost certainly I would learn that I have some variant of Marfan's syndrome that is not understood. Relatives who also have the same phenotype can get sequenced to verify that it is genetic. But since none of us are showing the most troublesome problems, and the literature almost certainly doesn't have our exact variation, I would learn nothing more than I know already. So just get an MRI every 5 years, and wait for costs to come down and research to improve. In 10-15 years look again and maybe the genetic test will be useful then. For a genetic condition as specific as mine, the sequencing done by these commercial providers is useless.
    - dekhn 8y ago
      I agree. I know a lot of genomics and health and after having my whole genome done, the results were worthless- I knew more than the genetic counsellors and had to explain to them that just randomly googling for SNPs and grabbinbg keywords from the research papers was a terrible idea. There are a few very well done genetic tests for a limited number of diseases.
      - Gatsky 8y ago
        Finally, a voice of reason.
- rootusrootus 8y ago
  Veritas does a full sequence for $999 and had a sale late last year for $200. I think Dante Labs will do it for $599. This is the whole thing, raw data included. Not 23andme crap. Edit: Dante seems to be doing the basic 30x full sequence for $229 at the moment.