Am I that predictable? Genetic variation and Jewish ancestry

Now I am not one to obsess about genealogy, at least not for my families. That is not to say that I am not interested in my family history, it is just that beyond a few generations (maybe 4-5 on one side), we can not go back very far, alas there is no paper trail.

This is not uncommon for people of Eastern European (Ashkenazi) Jewish heritage, like myself. Indeed, it is not uncommon for a lot of people from many different ethnic groups. Thus, not surprisingly, there are many people who wish they could know something about "where they came from", in a genealogical sense.

So where does the genetics come in ( this is a blog about genes)? Over the past 15 years there has been an explosion of studies that have used genetic variants in humans to help to try and identify the small proportion of the genome that differs between individuals, and use these variants to infer something about the ancestry of different people, and their ethnicity. This has been a fusion of questions from population genetics and anthropology.

What is population genetics you may ask? As the name implies it is the study of genetic variation within and between populations. It is a framework that we use to study evolutionary changes while considering variation within genes (alleles).

Population of what? Any organisms.. Including, and to most of you I may guess most importantly.. humans...

Now these studies of human population genetics have been given lots of press, and rightfully so. They are both extremely cool, and require all of us to think about what if any implications they may have. Of course the most important results have actually been known for along time. Most of the variation found in humans can be found within a population. That is, you do not need to go to some far flung tribe to find genetic diversity, you can probably go next door to your neighbors.

Now what does any of this have to do with the title of this post ("Am I predictable?"). Well there is a recent study by Anna Need et al. from the lab of David Goldstein entitled
"A genome-wide genetic signature of Jewish ancestry perfectly separates individuals
with and without full Jewish ancestry in a large random sample of European
Americans"

http://genomebiology.com/2009/10/1/R7

While the title may be mouth full, in all fairness I have to say my jaw dropped a bit when I read it. Why? Basically, the authors suggest that based on a very large (500 000) number of genetic markers that vary among people (genetic polymorphisms in the lingo), Jews of Eastern European ancestry (Ashkenazi Jews) can be perfectly identified as such, when compared to other people of European descent (we will come back to this last part at the end of the post).

Given that all of my grandparents were of Ashkenazi descent, this means that if they "genotyped" me (that is extracted my DNA and then checked me for all of these markers) they would be able to (almost) unequivically identify me as having Ashkenazi heritage. Of course the formal possibility exists that in fact my ancestors were not all Ashkenazi Jews, in which case, they could identify this as well. Of course, given that my physical appearance has been at one time or another been variously compared to David Schwimmer, Jeff Goldblum, David Krumholtz and Albert Einstein, I think the latter is a remote possibility.

Now, in the past few years there have been a number of studies that have suggested that the genetic ancestry of Jewish people is of Semitic descent. That is to say, Jews who hail from Europe appear to be more closely related to other Jewish people (whether from Europe, North Africa, Iraq or many of the other parts of the world that Jews settled and formed communities), rather than being genetically similar to their host communities. Thus despite a prolonged presence in places like Europe, North Africa and the middle east, Jewish populations have been maintained in large part by endogamy (which means Jews tended to have children with other Jews), and their has been relativey little local conversion to Judaism, providing an infusion of new genetic material.

Perhaps the most famous example of this has been the work on the markers on the Y chromosome. The Y chromosome is passed from father to son, and is remarkable in that bits of it do not undergo genetic recombination. This means, that regions of the Y chromosome do not get shuffled in the same way that other parts of the genome do. For reasons I will not go into right now, this provides an unusually informative set of genetic markers to help infer ancestry.

In a study a while back by Hammer et al., they examined 18 such markers on the Y chromosome for 29 populations including European, Arab, North African and Sub-Saharan African in addition to individuals from Jewish communities from all of these regions. What did they observe? Most of the jewish populations clustered with each other, and with non-Jewish groups of Arab descent (Syrian, Lebanese, Kurdish and Palestinian). Thus this evidence was consistent with Semetic ancestry for Jewish groups.

However, this (and other studies based on markers on the Y chromosome) only represent a small fraction of genetic polymorphisms (despite being quite informative), and more importantly only tell us about ancestry from father to son. We would like to make our conlcusions based on more than that right?

last year, two papers came out in the Journal PLOS Genetics
http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.0030236
&
http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.0040004

Where they mined large data sets collected originally for mapping genetic variants that contribute to disease susceptibility to examine the similarities and differences within and between populalations of people of European ancestry. Unlike the previous studies, that were based on a relatively small number of genetic markers, these studies used hundreds of thousands of polymorphic markers, spread throughout the genome.

These two studies found very similar observations with respect to European populations. First you could distinguish ancestry of people of European descent (i.e. people of Irish descent clustered with other people of Irish descent, people of German descent clustered with other people of German descent). However, the more interesting finding in my mind was that Ashkenazi Jews not only clustered together, but were quite distinct from other European groups. Not surprisingly, the European groups that Ashkenazi Jews were most similar to were the Mediterranean populations (Greeks, Italians and Spaniards). Indeed, I interpreted part of these results to suggest that Ashkenazi Jews were slightly more "Mediterranean ", than the other Mediterranean groups. However, given that this comparison was just among European groups, it does not in itself point to broader Semetic ancestry. However in combination with the previous results, the results are looking pretty solid.

So given all of these results, why was I so surprised by the paper by Anna Need and collegues? They suggest that not only do Ashkenazi Jews seperate out from other Europeans, but genetically can be very clearly distinguished, as I mentioned above. This is astonishing given that as a geneticist, I am used to thinking about probabilities. Normally when such genetic studies are performed, and indeed if you yourself get genotyped to find out about your ancestry, at best you will usually get an answer like "you have a 21% chance of being of X ancestry, 17% chance of being of Y ancestry). In other words, you seem to be a perfectly good human being, and anything else would be mostly guess work.

However, this study suggests that at least with respect to comparisons with Europeans, the probability of correctly identifying your ancestry to Ashkenazi Jewish ancestry is much higher. Is it going to really be a 100% accurate? No of course not, and the authors did not suggest this. Indeed they make the very careful point that is only in comparison to European populations. After adding information for Arab populations, it is clear that Ashkenazi Jews resemble some combination of European Mediterranean populations and near Eastern populations, and represent one more set of points along a continium as opposed to some discrete group. It is likely that as more Mediterranean and near eastern populations are examined (say people from Turkish, Armenian and Kurdish descent) there will be more overlap with these groups.

ok... I think I have said enough..

until next time