Readers of GNXP may remember a straw poll Razib took somewhat over a year ago asking which undersampled or unsampled populations people would like to see genotyped. The Ainu, who have heretofore really only been analyzed for Y- and mtDNA, were a frequent mention. As some of my previous posts probably made clear, the Ainu would have been my top choice, too.
Today, it looks like the wait’s up. As soon as the current issue of the Journal of Human Genetics (Volume 57, Issue 11) comes online, the world will see what 36 Ainu individuals look like at 900,000 SNP resolution. These are evidently coming from stored materials collected around 30 years ago. To accompany them will be 35 native Okinawan and, from Honshu and elsewhere — which hopefully includes Hokkaido — 243 Yamato Japanese samples.
Ancestry of all Japanese well-approximated by a two-way mix of “Yayoi” and “Jomon”; “Jomon” component bimodal, albeit still somewhat diluted relative to pre-Yayoi epochs, in far south (Okinawans) and far north (Ainu); Ainu closest, in decreasing order, to Okinawans, Yamato Japanese, Koreans, and Han Chinese; Kanto Japanese rather similar to Koreans, less so to Han Chinese (some of the articles suggest even closer to Koreans than to Okinawans, though that remains ambiguous).
None of this comes as a surprise. Physical-anthropological, archeological, and textual data (i.e., accounts of warfare with and enslavement of Ainu-like hostile emishi on Honshu as well as the more peaceable co-opting of friendly or pacified emishi into the Japanese world-system) have long indicated as much, and the last point in particular has been fairly undeniable since the 2010 Jung et al. study — no matter how much Japanese and Koreans alike have sidestepped it. (And continue to sidestep … look at what the Asahi Shimbun chose to highlight instead.)
It’s just the first time this story’s being told with an Affymetrix 6.0 array, and the first time we have any autosomal Ainu references whatsoever.
(Left from the Asahi Shimbun, right from the Chugoku Shimbun.)
It’s nice timing that this comes hot on the heels of the Loh et al. ALDER paper, which estimates the Japanese — using CHB Han Chinese as the Yayoi proxy — to have admixed 45 +/- 6 generations/1,300 years ago and a genome-wide lower bound “(likely very conservative)” of 41 +/- 3% “Yayoi” ancestry.