gene testing

I subscribed to the gene testing service from 23andMe some time back, and have been inspecting the results during the last couple of weeks now. Overall, I find the results interesting. I’m slightly concerned about the privacy and security aspects of this kind of a service. Not so much from the technical standpoint of this particular site, but from the nature of the data itself. I am also unsure if the ROI for this (169 €) is what I was looking for. I did, however, buy myself some peace of mind.

I’m no biologist, but more of an engineer, and don’t pretend to be an expert about this stuff. I’ll do my best not to butcher the facts. The whole process takes a month or so, including the two-way shipping of a saliva collection kit and the analysis of the spit with some kind of a chip technology.

genetic information #

The results list things like inherited conditions and genetic risk factors. The most sensitive ones like susceptibility to Alzheimer’s and Parkinson’s are not initially displayed, and must be unlocked after reading a few disclaimers. Apparently, I have basically no “bad variants” of the genes present for any of the studied conditions, and can evidently call myself lucky. The results were therefore somewhat boring, though in the best possible way.

Apparently, I have 3 % neanderthal DNA, which is slightly more than the average European. It’s actually in the 93rd percentile out of everyone, and I find that amusing. My genetic ancestry is from northern and eastern Europe, which is unsurprising. This is still all interesting, and there’s also a kind of a relative-matchmaking service that allows one to match with other genetically close users. The service hasn’t been available in Europe that long, and as I expected, I didn’t find any close matches.

Both health and ancestry data can be sensitive, so one should be aware of the possible bad news or surprises. I must admit my heart rate went up a few times, particularly when opening the “genetic risk factors” page.


Apparently, 23andMe had to pull the health-related genetic information because of the Food and Drug Administration a while back. Supposedly, FDA was concerned how this testing can be marketed, and how accurate it actually is. I don’t believe this would’ve ever been a problem for the European customers, but it highlights the fact that this stuff isn’t really an exact science. At least for a while.

data #

There’s a huge amount of data available from these tests, as the raw data can be viewed and downloaded anytime. The problem is that one wouldn’t know what to do with it. There’s a shitload of studies done on specific genes, but one should be careful when making conclusions based on something that specific. One trait can generally be affected by multiple genes, environment, diet, and so on.

I should suspect we can be very subjective about ourselves and be vulnerable to confirmation bias. The problems with big data I wondered about earlier, are no less applicable here.

security #

Collecting someone’s DNA is not difficult, but still requires some effort. But the fact that this data is collected wide-scale can be disconcerting, at least to those of use with more paranoid tendencies. The fact that my genetic information is now dumped in a database of an American company is somewhat unpleasant. One of 23andMe investors is Google, and it knows everything else about me already.

opportunities #

It’s not all bad, though. The user can agree for their data to be used in different studies, and that can of course yield positive results for themselves and others. Some of the studies conducted (or participated) by 23andMe are even shown on the “Traits” section of the service, and I find that fascinating.

I generally keep the more positive outlook on storing my information with some entity. I’d say I’m willing to trade a possible leak or mishandling of the data to the certain positive use I can achieve with it.

api and integrations #

23andMe provides an API, and all of a user’s information can be downloaded through it. The user must then extend their trust to these providers, too.

A few that I ran into are Dnafit, Athletigen and Promethease. Dnakit promises to create you a diet and exercise regimen based on your genetics. I’m yet to try it out, but it has been recommended by some health/fitness experts. Athletigen looks similar, but is free. I found it somewhat useful, though it’s still in beta and probably in an MVP, bare minimum stage.

Promethease, on the other hand, cost only a few bucks and basically provides and integration between 23andme and, which is like a database for studies linked to genes. It too, provides an overwhelming amount of information, and after about an hour or so of browsing through the studies linked to my particular genes I found my self no more wiser than when starting.

From what I understood, many of the provided studies are based on one gene. What’s confusing, is that other studies on same or different genes can provide the exact opposite effect. So one cannot base much on the effect of individual mutations, but would need a some kind of an accumulated report. I suspect that currently, this kind of data exists only for certain, more researched areas.

I wonder when we’ll see something like a Tinder or Pornhub integration, just because. Seriously, I’m sure there will be more useful integrations, thought it’ll take time. This data can’t be easily approachable by your average hacker, as it requires quite a lot of base knowledge to pull anything useful out of it.

is this good for anything #

I have the same problem with 23andMe, as I do with fitness trackers and the like. I get data from them, but little of it can be put to efficient or daily use. At best, they can infer and encourage positive behaviour. I must say though, that this data is a magnitude cooler and more interesting than the amount of steps I take.

What effects can this information have on healthcare, and perhaps unfairly, on insurance policies? Wearables are used at least by some healthcare companies, and what I’ve understood, insurance companies are also interested in the data. For now, I’m fairly healthy and the genetic tests revealed no risk factors or inherited problems whatsoever. I’m curious if that would grant me a discount on health insurance. Now that’s a small step towards all kinds of problems, philosophical and practical.

The best thing with this data is that it doesn’t change, and is likely to yield more facts and insights when time goes on. The test doesn’t cover the anyone’s whole genome, but parts of it. And what I’ve understood, 23andMe stores your saliva, so they can re-analyse it with a newer chip as technology advances. I’m intrigued by how this business will evolve.


Now read this

golang and positive constraints

I wrote a smallish program in Go around a year ago. Today, a colleague shared this article, and inspired by it, I went back and had a look at my now old project. The simplicity of it was surprising. an argument for any language # One can... Continue →