Biology and genetics are absolutely fascinating—if frustratingly complex—subjects, and although I’ve done quite a bit of research and studying on my own, I have neither the qualifications nor the experience of a professional; much of what I post here will very likely be oversimplified, misinterpreted, or just plain wrong.
Alright, so about a week ago, I mentioned I was going to write about my experience with 23andMe and what I found. Much procrastination, several revisions, and a frazzled “fuck it” later, I decided I may as well post something.
Because of a pending FDA investigation, 23andMe can’t provide health information at this time, but fortunately users can still download their raw data for analysis by a third-party tool. There are several available, but I chose two: Promethease and NutraHacker. NutraHacker is definitely a bit more hoakey of the two, providing (as the name suggests) recommendations for certain foods and nutrients to either “encourage” or “avoid.” I wasn’t able to see any sources cited for my report, but the recommendations seem to match up with the data from Promethease as well as my own observations over the years (though admittedly there’s a high potential for confirmation bias).
My Promethease report, on the other hand, provides everything a curious amateur biologist might want: information on which variations of which genes it found, conditions or traits they have been associated with, and SNPedia links (think Wikipedia, but tailored for Single Nucleotide Polymorphisms and genetics) with citations for everything. Immediately, I realized there’s so much I didn’t know I didn’t know about genetics and biology, but what I found most surprising was the number of genes with only a couple of citations at all—essentially the scientific equivalent of saying, “We have no freakin’ idea what this gene does.”
However, I’ve noticed the occasional SNP whose function and mechanism is moderately well-understood. For example, my report shows I have rs1065852(T;T), a variation of the gene for the GYP2D6 enzyme that metabolizes some medications. My particular genotype suggests I metabolize these medications relatively poorly (meaning the medications would remain in my bloodstream longer than in those with other variations of that gene, i.e. the therapeutic effect should last longer). This was of particular interest to me because of my difficulties over the years of finding the right medications that work for me; I’m excited to see where genetic research leads in the coming years specifically in relation to pharmaceuticals.
It can be very tempting to look at something like Promethease’s report and misinterpret it. In fact, that’s a key argument in FDA’s suit against 23andMe: that the general public does not know how to properly interpret these analyses. Scientific and statistical literacy are topics that, despite being vital to our understanding of the world around us, are either poorly taught or difficult for many to grasp.
With this in mind and a healthy bit of prudent scientific skepticism, I still can’t help but notice some things that stood out to me. For example, rs3803662 is related to an increased risk in breast cancer (primarily with women, obviously), but I wonder if that played any role in my mom’s developing breast cancer a few years ago (since one of my copies of that gene is from my mother). Could rs5186 (“~1.4x increased risk of hypertension”), rs7961152 (“1.5x higher risk for hypertension”), rs4961 (“1.8x increased risk for high blood pressure”), rs6997709 (“1.5x higher risk for hypertension”), etc have been why I developed high blood pressure at a young age? Maybe my gs227 (“fully heterozygous, age related tasting variation”) explains why I’m moderately better at cooking than the average person.
Who knows? But I’m glad I did this. If nothing else, I’m glad I’ve been able to contribute a few more data points in order to help scientists figure out the enigma that is the human genome.