Freitag, 25. Oktober 2013

Following the own long distance mtdna branch

You need patience to focus on your own mtdna lineage, because mutations don't happen often. But it is a bit like staring on new fotos - all people are looking for themselves and I'm no exception.
For more than two years there had been no news about my haplogroup K1b2a and Mannis van Oven, the master of phylotree, mailed me in october 2011: "You are a very close match with GenBank accession EU621708, which is in K1b2a and also has 12188C". The last release of phylotree (built 15 in sept 2012) went by and there were no further samples with my mutation, so new subclades like K1b2a2 were given to others. As an artist you are always a bit ahead of your time and I regarded myself as K1b2a3, although it wasn't confirmed.
But today I found in the supplementary files of the new publication of Marta Costa both my private mutation 12188C and the confirmation of my subclade guessing.


  1. I'm quite surprised of the old age that you attribute to haplogroup K in general and its derived branches. K is first documented in the aDNA of Syria (Middle Euphrates) and looks one of the most favored lineages by Neolithic and post-Neolithic expansion, it's several star-like substructures (incl. K1a1b, K1a4a, K1c1 and K2a) would seem to correlate to such Neolithic dynamism. I did in the past ponder if K might have expanded with Magdalenian but I have not found any evidence in support of it. My own incursions into "molecular clock" approximates strongly suggest that K in general is very recent, although it's difficult to estimate a realistic age on that alone.

    I can agree on the age estimates you make for U and U8 but downstream of that I feel that it does not make much sense. Let me explain: U and U8 are separated by a single coding region (c.r. hereafter) mutation, while U8 and U8b are separated by three of them, therefore the logic suggests that the time to coalesce should be quite larger for U8b (triple in principle). Same for the U8b-K stem, which also has three c.r. transitions, etc. So, following this logic, I'd place U8b at ~35 Ka and K at ~20 Ka BP (more or less). K1 could well be then from c. 10 Ka BP, what places it just in the right time to be found in the Neolithic expansion process.

    You'd ask: what happens with the more recent mutations in downstream branches? In Paleolithic conditions (very low population densities, very slow demographic growth, if at all) most of them would not survive drift, so it's a "modern" phenomenon, proper of generally expansive populations, that so many of them survive (or even any of them at all). However it's possible that they may still weight for some adjustments towards older dates.

    In any case your particular lineage K1b2a3 seems restricted, from your reported data, to Germany and France (maybe some other nearby region), so it's very plausible that it had a quite recent expansion date. Your estimate of c. 3000 Ka BP would be coincident with Celtic expansions and such, however I would not be surprised if it has a much more recent date, like in the context of the Germanic migrations and the Frankish Empire (half your estimate or less).

    In any case, congratulations for your newly acquired ancestry knowledge.

    1. thanks Maju for reflecting this. I know your concerns about a molecular clock and generally agree. Mistrusting any ratio based on a certain interval is quite reasonable, I think, since we cannot be sure, wether a second mutation occured 50 or 5000 years later. Until we have confirmed data by ancient dna, all is left to speculation and I have no way to proove my guess.

      I added an older pic about the mutational way from CRS to my subclade, which is consistent with your intervals. Only your interpretation is different. If you agree about the age of R and U, than three CR mutations are somewhat like 5000 years, so three CR mutations plus 2 HVR1 mutations estimating as 10000 years, as I did between U8b and K, for me is quite reasonable. The three oldest subclades of U are for me U5, U8 and U2, which can be placed between 48 and 52 ky bp. So U8b can be around 42 - 45 ky and estimating the next step of K with an caucasian origin somewhat 10 ky later for me is logic. Brian Sykes first guess of 16 ky in north Italy is rather interferring newer publications. I like the estimations of the new Costa paper, which not only confirms my personal guess, but also supports my interpretation about the arrival of early Jewish people in Germany. Furthermore we will have to learn more about several waves of K-subclades immigrating Europe. Some of the Syrian ones, you cited seem to be specific Near-East, while others derived already in Europe. We seem to be just at the beginning of an understanding process about different immigration waves.