23andMe

[Anonymous No.8196932]

So I bought 23andMe and I'm unsure whether I should do it.

Is it dangerous?

Is it just a complete botnet ploy by Google?

Serious answers, should I return it?

[Anonymous No.8196941]

don't do it if you don't know how to read it to see how truly retarded and flawed it is

it's a ploy to strip people of their money

[Anonymous No.8196944]

>>8196941
What about it is unreliable?

[Anonymous No.8196949]

>>8196944
a lot of their studies
for example they will tell you that you don't have a breast cancer allele or risk factor based on a study done on ashkenazi jews - a group which statistically you probably don't belong to

a lot of the studies to back up their stuff are not optimal and you have to do some digging to find that out

also they just don't account for how different populations are, for how few studies there are done on some gene etc.

they might end up scaring you or depressing you or giving you a false sense of security because of some bullshit that actually isn't true and you've wasted your money

[Anonymous No.8197200]

>>8196949
You can still check whether you're at risk on other websites. As far as I know, their gene sequencing is reliable.

[Anonymous No.8197408]

>is it dangerous?
not at all, you spit in a tube and put it in the mail
>is it a botnet ploy
absolutely. they're not going to use the information themselves, but they're collecting an enormous database of phenotype information paired with genotype information that they can monetize for insurance companies and the like

let's take a step back and look at the major advances in agriculture in the last couple decades, specifically in dairy production. if you look at, say, milk production, over the last couple decades we've drastically increased the volume of milk we produce per cow without increasing our land usage. that was possible because of genetic SNP detection and statistical inference methods that allowed us to make incredibly effective crosses between cows. cheap SNP detection paired with enormous breeding pools let us infer the animals with the best combination of predicted presence of positive traits and lack of negative traits

now let's say we wanted to do a similar thing with humans for some trait, say, detect which people are most likely to develop cancer or heart disease for insurance purposes. we'd need a LOT of information on pedigrees, on phenotypic information, and SNP information. once you have that, the math is fairly well worked out. the chips at 23andme aren't quite loss leaders (microarray assays are pretty damn cheap, especially in bulk volume), but they're not the real product that's going to make the company rich