Gene testing; neat things and some stats jackassery

I  got myself gene tested for the lulz and the raw data.  I was primarily interested in ancestry results. For my ancestry results in the broad strokes, check out the article I recently wrote for Takimag.

Over the years, quite a few physical anthropologists have learned interesting things about historical human migrations from examining genetic data. I was hoping to learn what tribes my ancestors hail from. I was pretty sure they’d be Celts, Germans and/or paleolithic non-Indo Europeans like the Basques (who appear to make up a large fraction of the European population). The results indicate that, in fact, that’s what my genes and chromosomes are made of. I can’t tell how much of each, as the science on this subject isn’t really done yet. Where it is done, the databases are not yet publicly available for comparisons. There are interesting tools for getting more detail than 23andme.com and FTDNA give on your genetic ancestry (both services allow the individual to download their results). The one I liked the best was the interpretome. The interpretome has a tool for doing PCA on reference populations. The results put me closest to the Swiss in my genetic makeup among European populations. This makes a lot of sense, as I’m part French, part British (aka Germano Celtic) and part Bavarian German.

 

 

The results of doing PCA against non-European populations are interesting, though they mostly confirm what is already known about human diversity. My African results? Well, Africa is a very genetically diverse continent; it shows me as closest to a Berber tribe called the Mozabite.

 

 

My Asian PCA results: again, people think of Asia as China, India and Japan. There are also other tribes. The tribe I ended up closest to? The Kalash people, who are related to the pagan Nuristanis of “Man Who Would Be King” fame. They’re not descendants of Alexander the Great’s Army after all (they seem culturally more related to old Persians), but they look like they could be.

For “medical” results: my test results indicated I got the luck of the draw in the brains and self-discipline department. They also indicated that I also have no genetic potential for body building or sprinting, something which is painfully obvious to me whenever I go to the gym. My genes peg me for asthma and hay fever, which is true. They also nailed me as brown eyed, lactose tolerant and sensitive to caffeine. On the other hand, the fancy pants genetic testing got my blood type wrong, which seems rather odd.

Gene testing is feared for its ability to classify people by their genetic potential. Looking at the results: I don’t understand people’s fears. A quick look at my CV, a quick look at me, and feeding me a fancy coffee drink in the presence of a cat would tell you all of these things in a few minutes: no affymetrix chips required.

There are some weak genetic correlates with common diseases. The gene tests tell me I have a higher probability of glaucoma than average, and a lower probability of diabetes than average. This is what I would expect from looking at my immediate family’s medical history. I’m also supposedly immune to norovirus infection, which could help explain my cast iron stomach. On the other hand, my gene tests indicate I have both a higher and lower chance of celiac disease and Parkinsons disease. The science behind these results is obviously inconclusive at best.  Possibly, it is complete nonsense.

Many of the studies the genetic tests are based on were conducted on a mere handful of people who I may have nothing in common with. Quite a few of these genetic studies of health outcomes do not pass the statistical sniff test and are likely to be measurements of random noise; caveat emptor. 23andme generally provides links to the studies of interest. I found this tool, by Enlis Genomics to be even more useful for sating my curiosity.

Is any of this useful, or is this just a high-tech version of astrology? The genealogical data is vague and often questionable, but useful to the curious. While I have no immediate plans for breeding, knowing that I carry no genes for purely genetic diseases such as Cystic Fibrosis is reassuring, though I suspected as much from family history. If you don’t know your family history, or have a very small family, such tests could be very helpful. I don’t think the “abilities” tests are useful at all; most of it can be told by looking in the mirror. Finally, virtually all of the disease risk and other tests should presently be taken with an enormous grain of salt.

About these ads