Search billions of records on


A new scientific paper has been released entitled “High-throughput sequencing of complete human mtDNA genomes from the Caucasus and West Asia: high diversity and demographic inferences” by Schönberg et al. In conjunction with the paper, 147 full mtDNA sequences (in Supplementary Table 1) from Armenia, Azerbaijan, Georgia, Iran and Turkey were added to the GenBank database. Of those, seven are clearly in haplogroup K (two in K1a and one each in K, K1a4, K1a4a1, K1b1 and K1b2a) and one in U8b. Two others are probably in K, but due to problems probably related to the “high-throughput” method to produce the results, are shown as being in haplogroup R0. (Others have reported problems with sequences in this paper.)

If you remember, Brian Sykes in his “The Seven Daughters of Eve” suggested an origin for “Katrine” in Italy. Others, including Dr. Doron Behar in his paper with the K tree which we use, have found K subclades originating in the Near East (aka Middle East). The Google map on the K Project website only has pins in Turkey, although there are a lot of those and some are ethnic Armenians. Of special note is the K1a4 in Turkey, but close to Georgia and Armenia. (The pin in North Ossetia in Russia, just over the line from Georgia, may be ignored for this purpose, since it represents a branch of K2a only otherwise found in the British Isles.) However, there are two K Project members with full-sequence (FGS) results tracing back to Armenia, in K1a and K1a4; one tracing to Iran, a K1a; and five tracing to Turkey, in K1a, K1a1b, two in K1a4, and one in K1a4b.

Comparing the new paper’s sequences with the ones in the K Project, K1a and K1a4 are found in both. I suspect that the K1a4a1 and the K1b sequences may represent back migrations. K1c examples are not found in either set from these countries, and I only see one K2a in Turkey on our map.

Let me note that the current version of the PhyloTree has K right below U8b, instead of U8b’K as still used by FTDNA. The U8b with the current paper is from Armenia; it’s actually in a deep branch of U8b1 alongside two “sister” sequences, one from a paper by Achilli and one from FTDNA – both of which list maternal origins in Italy. What has worried me in the past is how K could possibly have originated in the Caucasus if no examples of its parent U8b were found there. Now we have one such example. (Do note that there is one U8b from Jordan in the Gonzales paper, but all others are from Europe.)

Now look at the unique “plain” K with the new paper. It is not a K1. It is not a K2. Assuming it is not beset with the difficulties resulting from the high-throughput method of producing its mutation list, it perhaps should be called a K* (a K not in a lower subclade), although FTDNA rarely, if ever, uses the * in haplogroup or subclade designations. (Being a singleton example, it can’t be called a K3.)

Finally, the new U8b was collected in Erevan (or Yerevan), Armenia, while the new K* was collected in Batumi, Georgia. Those two locations are just 180 miles apart, with the Armenia-Georgia-Turkey common point about half-way between. So, I still may not know where K originated; but I have a new place to look!

Bill Hurst

Administrator, mtDNA Haplogroups K and U8