Hacker News new | past | comments | ask | show | jobs | submit login
How To Use 23andMe Irresponsibly (2014) (slatestarcodex.com)
119 points by danielfoster on May 6, 2017 | hide | past | favorite | 40 comments



I don't know if it was around in 2014, but https://promethease.com/ can do this analysis for you, including from AncestryDNA and 23andme. I plan on using it once my AncestryDNA results finally come back.


I used promethease and learned I'm CCR5-Δ32 and likely HIV immune

https://en.m.wikipedia.org/wiki/CCR5


also -- https://dna.land/ seems to give extremely similar (possibly identical?) results, for free


Neat! Might as well do both and see how they compare :)


I just plugged my 23andme data into both. Waiting for results.

I was thinking of sharing the results, is anyone interested? Is there really any good reason not to? I know people get all uptight about it but I'm not sure why.


I did this [1], and so have thousands of others through the Harvard Personal Genome Project and other associated projects. So far, I haven't faced any consequences. If you decide to do so, I encourage you to make your genes accessible to researchers by participating in the Harvard Personal Genome Project [2].

The main risk of sharing your genes is genetic discrimination. It's irreversible to share your genes, so I'd ask yourself: if you were to find out you are a carrier of a genetic disorder, would that change your mind? If so, I wouldn't share your genes publically.

[1] http://genes.howderek.com/ [2] http://personalgenomes.org/


Promethease report is interesting...

https://adamjacobmuller.github.io/promethease.html

I'll do the PGP thing at some point... I get they want informed consent before I do it but its now asking me to review pages of material and making me take a test first...


Results I dont think anyone gets uptight about, the actual DNA data absolutely.

People get uptight because in the future someone like an employment agency or life insurance company might buy one of these services and use the data to discriminate.

There's also social consequences that may come of this, like if you had a one night stand, public publishing your DNA might create a bread crumb for your child to find you which could be a good or bad thing depending on if you had a spouse at the time for example.


I found out I have a surprise half-sister through 23andme. It also turns out that my Dad may not actually be my biological father.

It affects you more than you might think.

I personally want the truth, no matter what, so after a few weeks of thinking about things, I was ok with what I learned. The hard part is dealing with family members than may not feel the same. People need to be aware of just how messy this can get.


> surprise half-sister

> Dad may not actually be my biological father

Wow, what a pair of bombshells!

Was this stuff known to both your parents or did it blow someone's cover?


Unfortunately, I can't ask them, as my mother has passed away, and my father will not talk about it. The fact that my Dad wasn't interested in genetic testing was my first hint, as that is very unlike him. But it makes sense in the larger pattern- I am the youngest of the family by 15(!) years, and all of my siblings were adopted. It seems pretty clear in retrospect that my parents likely had to use a sperm bank. I've had my suspicions for a while, so this wasn't the complete shock it could have been.

Still threw me for a loop at first, though.


Oh right, a sperm bank would explain both factors neatly. Certainly loop-worthy though!


How did you find out about your half sister? I thought 23&me only gives you information about yourself and nobody else?!


She contacted me via 23andMe. I could have left it at that, they don't give out details. And I had to chose to participate in the 'find your relatives' feature. It's as open or closed as you want it to be.


The Holocaust happened fewer than 100 years ago. You don't want to publish such identifying data with humanity still so immature.

It could turn out that you have the "dissident" gene and the government needs to get rid of you to preemptively stop the problems you might cause.


It's funny how different people have different approaches towards dealing with such a potential scenario.

For instance, to some people, preventing others from accessing this information is a mechanism to prevent such a thing from happening in future.

For others, having gun rights or living in a heavily armed society (such as USA) is a better preventative mechanism.

FYI in N Korea dissidents 'genes' are removed from the society, but it's usually done by taking out all the family members of a person committing crime. So I'd say that if govt really wanted to do something, this won't really be a defense.


The 2nd amendment existed in 1942, but that didn't prevent the Japanese from being rounded up. We didn't decide to murder them, but we could have.


i don't have a concrete reason besides personal preference, but it seems like the kind of thing that might become very regrettable for unforeseeable reasons in a decade or two so i would advise against it


I also wondered why people are so freaked out by it, asked around, and came up with these more or less worrying reasons:

http://gleech.org/genes-out


23andme currently offer an Ancestry Service, and a Health + Ancestry Service. Are you using just the Ancestry service here? Anyone know how their Health reports stack up against these 3rd party services?


I did 23AndMe a long time ago, before they had been neutered by the FDA. I wouldn't use them again for that reason.


dna.land is very nice, they do trait surveys and compare it to their predictions so you can actively help improve their data.


I don't think Mr. Alexander has had this conversation with anyone in bioinformatics, where "SNP" is universally pronounced "snip" and given the preconsonantal indefinite article - "a snip", not "an ess en pee".


Why is this an issue? It's not like he's trying to make a career out of it.


It fell oddly on my mind's ear and so seemed worth mentioning. Perhaps it wasn't.


Like a lot of stuff that comes out of the rationalists -- autodidacts with a background in stats have decided that reading research papers on their own is a sufficient approach to knowledge, with often pretty messy results.


man what a refined piece of pretentious snobbery - finding fault with someone who reads research papers as a hobby. ha.


His actual reading appears to be spot on, which ia fairly impressive, and well nuanced besides with regard to reliability of results in a field not especially meriting respect for its rigor.


Has anyone used 23andme with fake information? If so, how?


Tell a work friend or someone you know causally about 23andme. If they show interest suggest buying the kits together to save on shipping costs. Mention that you have a few friends who would be interested as well and give them the cash for a few kits. Their credit card and address will be linked to the purchase. When the kits arrive sell your extras to different causal friends and register yours with completely fake information from public wifi. 23andme will have your DNA but nothing to link it to your identity.


Why the extras?


I'd assume it's to prevent your friend from being able to know that the extra kit was yours? Though if all your friends use their correct information - that information is still leaked, hence the need to distribute them to friends outside of the same social circle to minimise the chances.


With a little cross referencing of third party "anonymized" metadata such as cellphone location, or facebook graph (works with other data set) it is very easy to uncover all the identities involved in this scenario with a high probability of success (>90%).

You'd have to push preventive measures further than distributing outside of the same social circle if you are serious about this.

One should also consider the legal aspect of this as there are places where faking your online identity is illegal.


Here's a pretty good article on exactly this question: https://www.abine.com/blog/2013/23andme-without-violating-yo...


Interesting read but too bad it is not practical for people don't work at albine (company making online privacy tools) and falls short of preventing your DNA code from being added to the large collection that is or will be used for whatever unknown purposes.


In Australia you can get pre-paid credit cards from Australia Post and use that for purchase.


You could take a strangers saliva


I have submitted a swab sample (for $100) with https://bethematch.org

I have emailed them to see if I can get a text file or be able to share my info with https://dna.land


https://you.23andme.com/tools/data lets you search your raw gene data for specific genes and markers(SNPs). If you're curious about having specific gene variants from this article, just search it on this page.


[ 2014 ]




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: