Category Archives: Family history

Untangling my genetic heritage, part I: Slav and Avars?

Getting my results from 23andme was only the first step. The main reason I did this was to probe my ancestry, but 23andme actually offers only a limited amount of information in that regard. But there was still some really good stuff. First: my mitochondrial haplotype. Mitochondrial genes are inherited directly along the maternal line,* without recombination, so it’s a great tool for testing ‘deep ancestry’. Mitochondrial genomes only change through the accumulation of mutations, and since this happens fairly slowly, it’s possible to infer human migrations over the course of thousands of years.

My mother’s family is German – from Ostfriesland on my grandfather’s side, and Sachsen-Anhalt on my grandmother’s. Until recently it never crossed my mind that they could be anything but German – family trees, going back to the 1700s in some cases, showed nothing unusual. Then a couple years ago (is it that long already!) I received substantial additions to my grandmother’s mother’s family, going back in some cases to the 1600s. Among those records were two non-German surnames – a Gorgas, b. 1753 and a Stoi, b. 1740. After Google proved supremely unhelpful in trying to figure out associate these names with various eastern European ethnicities, I tried the Ellis Island database on immigrants to the US. While Gorgas seemed to be found everywhere from Greece to Lithuania, the Stoi I found was Romanian. The only connection to Romania I could possibly come up with was the fact that German miners had been in demand in eastern Europe, and my grandmother’s ancestors had been mining engineers. Maybe someone had migrated to a Transylvanian Saxon settlement, found a wife, and returned back home. Far-fetched, yes, but it was the best narrative I could come up with.

About a month and a half ago, I stumbled upon a better narrative. Razib posted an article about Sorbs, relics of the Ostsiedlung, he called them. After the westward migration of Germanic tribes into the lands of Roman Empire (the “Barbarian Invasions”), Slavic tribes expanded into the lands between the Elbe and Oder rivers. The Carolingians expanded eastward, incorporating not only the pagan Saxons, but also a wide swath of Slavs and Avars into their empire. Thus began the Germanisation of many of these Slavic groups. The town my grandmother came from, Großörner, was located just west of the Slavic lands. Since women are less likely to move than men, it seemed possible that my mitochondrial genome showed Slavic roots.

I approached my 23andme results hoping for a mitchondrial haplogroup that was more common in eastern Europe, something that might point to possible Slavic ancestry. But when I clicked on those results, I was truly baffled. The graphic showed a mitochondrial family that was centred India. After a bit of puzzlement, I realised that the graphic showed results for haplogroups U2 as a whole, not the specifc U2d hapologroup. I recorded my initial reaction in a comment at Gene Expression.

Just got my 23andme results a couple days ago, and I took a moment to think about my reaction to them. The first surprise came from my mt haplogroup – U2d. First glance at the distribution map on 23andme was really confusing, since it showed the entire U clade, which was predominantly Indian. Since I’m Indian on my father’s side, not my mother’s, it was puzzling. A bit more reading left me, if anything, more confused, since they said that U2d was primarily Jordanian and Palestinian. I considered for a moment whether this meant Jewish ancestry, but then I realised that if they meant Jewish, they would have said Jewish.

So off to Google Scholar, which turned up this paper. That discovery tied things together in a way that was even better than I had hoped. Turns out that U2d is found and parts of the Caucasus, and among Czechs in western Bohemia, together with a relatively high frequency of east Eurasian mt lineages. The authors (in a previous paper) suggested that this may reflect Hun or Avar incursions. Why I find this all so cool is that after reading Razib’s blog post about the Sorbs, it crossed my mind that I might find a Slavic signal in my mt DNA, given my maternal lineage’s deep roots in Sachsen-Anhalt. Given the whole “men move, women stay put” idea, it seemed reasonable to posit Slavic ancestry. Finding what could possibly be Avar or Hunnic ancestry though, is even more fun.

Genes and ancestry

Back in April, 23andme had a special offer by which you could get their personal genome screen for just the price of a one-year membership. Having read Gene Expression off and on for the last few years, I have gradually grown more interested in the world of personal genomics, but what really caught my attention was when Zack Ajmal started his Harappa Ancestry Project back in January. His focus on Indian ancestry really opened up new possibilities for me, and by following his blog (and paying more attention to Razib’s posts on the subject) I felt a great to desire to join in the fun.

So, back in April, I ordered my kit. And then I did nothing with it for a few weeks. It was nice to contemplate it, savour the sense of possibility. But there was also the matter of not eating, drinking or brushing your teeth within half an hour of giving the (saliva) sample. And it seemed like every time I realise that the kit was sitting there – I had just broken one of those rules. Eventually I got myself to do it, mailed off the kit, and a few days ago I finally got my results. And I was faced with a new question – now what?

The first set of results I got had to do with things like disease status and traits. I learned that I am likely to be at higher risk for some diseases, at lower risk for others. It was interesting observe my own reactions to these results – my initial reaction was to downplay both the negatives (the higher risks of certain diseases) while also downplaying the positives. The former makes sense – the odds of most of these things are low enough that even if my risk is higher, it’s still very small. On the other hand, even if my risk of Type II diabetes is about 20% lower than the average, it’s still shockingly high – mostly because the average risk is shockingly high. After a while though, I came back to the realisation that, with regards to the things that might pose a higher risk…well, I need to use that information to my advantage, keep an eye out for things (without being a hypochondriac) and maybe a little knowledge might let me catch a problem sooner, rather than later… (More later)