Dienekes points me to a provisional open access paper, Analysis of genetic variation in Ashkenazi Jews by high density SNP genotyping. Here's the conclusion:
There were small but significant differences in measures of genetic diversity between
AJ [Ashkenazi Jewish] and CEU [Utah whites from the HapMap sample]. Analysis of genome-wide LD structure revealed a greater number of haplotype blocks which tended to be smaller in AJ. There was essentially no difference in global LD decay between AJ and CEU, although there was a tendency for faster decay of nearby SNPs and slower decay of intermediate distance SNPs in the AJ. These data are more consistent with the AJ as an older, larger population than CEU, and would suggest that, depending on regional differences in LD structure, AJ populations may not always provide an advantage for whole-genome association mapping.
The image above left illustrates the differentiation between 102 Ashkanezi Jews and 60 gentile Northern Europeans (a Mormon sample from Utah) along the first two principle components of genetic variation. Obviously there are average genetic differences between gentiles and Jews, as evidenced by the different prevalence of some diseases and so forth. I am intrigued by the result of a larger older population, though I'm not sure if I want to read too much into it. Do remember though that long term effective population is the harmonic mean, which means that one generation of an incredible bottleneck can have a disproportionate impact. For more of the same, see Discerning the Ancestry of European Americans in Genetic Association Studies.
- Log in to post comments
Re: the population sizes, would a more reasonable non-mathematical rendering be that the smallest AJ bottleneck was larger than the smallest CEU bottleneck, or the average AJ population size was larger than the average CEU population size?
Re Dan's question: either answer could be right. Instead of thinking about effective size think of the inverse of effective size, (1/Ne), which is the rate of diversity loss. This inverse can be averaged over time, like any speed. A bottleneck has a much larger effect on the average of (1/Ne) than it does on average Ne.
Will this put an end to the talk about a bottleneck in Ashkenazi history and about Ashkenazi disorders being the result of drift?
Henry
I've been in an argument here over whether Ashkenazi jews cluster closer to Northern Europeans or Arabs. I did some searching at this site, but I wondered if you could point me to something more conclusive.
I've been in an argument here over whether Ashkenazi jews cluster closer to Northern Europeans or Arabs. I did some searching at this site, but I wondered if you could point me to something more conclusive.
greg cochran told me that henry had run a few hundred SNPs and they were closer to northern europs than yemeni jews (ashkenazis that is). i haven't seen any results that include arabs though.... (i think it would be a close thing).
Finding high inter-group separation between modern American descendents of NW Europeans and Ashkenazi Jews, with all the confounding factors at work (physical separation of the original lineages + bottleneck effects imposed by emigration to the US) is cool, but not really surprising. Especially with hundreds of thousands of SNPs (I'd bet you could distinguish between UK cities at such a resolution).
The really interesting thing would be to compare European Ashkenazi Jews with their historical neighbours - i.e. Slavs and Southern Germans. As in, take modern Jewish Poles and non-Jewish Poles from the same region, and see how much separation you get between the clusters. This would give clear information about the degree of separation between Jews and non-Jews.
bottleneck effects imposed by emigration to the US
Bottleneck refers to when the population gets really small in absolute terms, and there's a sudden great loss in diversity, which hits the genome indiscriminantly. Nothing like that happened.
Bottleneck refers to when the population gets really small in absolute terms, and there's a sudden great loss in diversity, which hits the genome indiscriminantly.
Well, considering that only a small portion (less than 10%) of the historical lineage of Ashkenazi Jews ever made it to the USA (compared to the sum total of those who migrated to Israel, stayed in Europe, and died in the Holocaust), I'm not sure how this doesn't fit your description... Same thing for NW Europeans, minus the Holocaust bit.
One might argue that the sheer numbers of migrants make the effect irrelevant, but I'd be inclined to think that you will find more variance among modern NW European than among US "pure blood" descendents of NW Europeans. If the same is true for the Ashkenazis, and since PCA is based on total variance, then it seems that lower intra-group variance will mechanically "magnify" the inter-group variance on the graphs (in comparison to what it would have been if applied to the original lineages). Or am I missing something?
Well, considering that only a small portion (less than 10%) of the historical lineage of Ashkenazi Jews ever made it to the USA (compared to the sum total of those who migrated to Israel, stayed in Europe, and died in the Holocaust), I'm not sure how this doesn't fit your description... Same thing for NW Europeans, minus the Holocaust bit.
on the order of 1 million people emigrated. they weren't that geographically skewed either in a unrepresentative manner (i.e., they came from galicia, they came from lithuania, they came from germany, etc.). so i think you're wrong. which is what this paper implies too (no bottleneck).
One might argue that the sheer numbers of migrants make the effect irrelevant, but I'd be inclined to think that you will find more variance among modern NW European than among US "pure blood" descendents of NW Europeans.
they address this in the paper and say it seems unlikely seeing as how the polygynous mormon descendants in utah show the same frequencies of NW europeans. the polygynous mormon sample is probably the best case for bottleneck effect, so if you don't see it there... rather than bottleneck, i would look to selection to be more salient for descendants of early american settlers; they bred like rabbits (e.g., more than 10 births per woman in early mass.).