23andMe

I didn't want to horn in on Dave's dna quest for history, but I had something neat happen in a similar vein.

So, first some background: My dad and his side of the family all suck. He was a drunkard and an abuser. His mom was cut from the same cloth. When my mom divorced him some 40+ years ago, he pretty much dropped out of our lives. So I know fuck-all about that side of the family.

So, one of my cousins from that side of the family (my uncle's daughter) just did 23andMe recently, and I popped up on her list. She never even knew we existed, because the aforementioned grandmother had told her mom that we had all died in a car accident (real winner there). So we've just spent the last several hours catching up on family business and commiserating on how much that side of the family blows. It was neat. I never expected to find random relatives on the service.
 
The beautiful part is everyone paid to be sold; the elegant part is that if a close relative of yours did so, your milkshake is more or less drunk too. All in all, this is solid "everyone saw it coming" news, but it's useful to remember that all the quips and jokes and predictions [did] indeed [end] in "it coming"
--Patrick
 
I have not seen the terms under which DNA access was granted, nor do I know their retention policies, so I cannot say how easy it is to reverse-engineer any “John Doe” back into the original PII.

—Patrick
 
23andme is a genotyping service so they don't have your full genome, unless they have kept some DNA frozen during processing.

I assume 23andme participants answered a lengthy questionnaire about demographics and health backgrounds? Can this data be trusted to be accurate?

These pharmaceutical companies could make $$$ off of the data. That would annoy me, but I guess Google et al. Is already doing a bang-up job of that.

I need to go watch gattaca again.
 
23andme is a genotyping service so they don't have your full genome, unless they have kept some DNA frozen during processing.

I assume 23andme participants answered a lengthy questionnaire about demographics and health backgrounds? Can this data be trusted to be accurate?

These pharmaceutical companies could make $$$ off of the data. That would annoy me, but I guess Google et al. Is already doing a bang-up job of that.

I need to go watch gattaca again.
There was no questionnaire when I submitted my sample: you spit in a tube and mail it back, and they give you results on the website. If you signed up for the Health info, you can participate in research surveys (which are usually just a couple questions at a time) about your health and life habits, but that is entirely voluntary. There were no questionnaires about ethnicity or cultural background.
 
There was no questionnaire when I submitted my sample: you spit in a tube and mail it back, and they give you results on the website. If you signed up for the Health info, you can participate in research surveys (which are usually just a couple questions at a time) about your health and life habits, but that is entirely voluntary. There were no questionnaires about ethnicity or cultural background.
Interesting. I wonder what these pharma bros are going to use the data for.
 
I wonder what these pharma bros are going to use the data for.
Most likely as input for machine learning models to study genetic pharmacology, or "Pharmacogenomics," which is essentially the study of just how strongly specific genomes affect drug responses. Science has already confirmed that a person's genetics can strongly influence their responses--redheaded people have higher pain tolerance, but they also tend to require higher/more frequent dosage of anesthetics, for example. To them, 23&Me is a VAST treasure trove of potentially relevant data, but it's more data than a team of human researchers could easily sift through, so they probably want to feed it all into a digital sieve of some sort to be a kind of robotic coin-sorter, but for DNA.

--Patrick
 
Top