Admixture test: Lithuanian mystery or unreliability of results

I made recently some f4-statistics in purpose to find out the Scandinavian admixture among East Europeans and Finns.  My formula was

f4(test,West-Finland;Norwegian,Mbuti)

The result made me wondering how much the genetic structure of sample groups has effect on the result:

Lithuania West-Finland  :     Norway      MBUTI      0.0027     2.905
Poland West-Finland  :     Norway      MBUTI     -0.0004    -0.583
Belarussia West-Finland  :     Norway      MBUTI     -0.0013    -1.620
RU_Smolensk West-Finland  :     Norway      MBUTI     -0.0002    -0.155
RU_west West-Finland  :     Norway      MBUTI     -0.0030    -3.779
RU_Tver West-Finland  :     Norway      MBUTI     -0.0031    -2.589
RU-Kostr West-Finland  :     Norway      MBUTI     -0.0054    -5.131
Romania West-Finland  :     Norway      MBUTI     -0.0105   -10.318
Bulgaria West-Finland  :     Norway      MBUTI     -0.0094   -10.234
Ukraine West-Finland  :     Norway      MBUTI     -0.0033    -4.404
Chuvash West-Finland  :     Norway      MBUTI     -0.0231   -24.172

Chuvashes are for a control purpose.


The result shows that the Lithuanians show more common admixture with Norwegians than the Finns do and they are the only group exceeding Western Finnish - Norwegian admixture.
 
Lithuanians share also more admixture with Chuvashes than Mordvas and North-Russians do.  They show more admixture than their neighbours, as well in east as in west. 

Mordva  Lithuania  :    Chuvash      MBUTI     -0.0039    -4.050
RU_Vologda  Lithuania  :    Chuvash      MBUTI     -0.0033    -3.406

Then I replaced the data base to the one published by Lazaridis et al. 2014, to confirm this observation.  I also changed the outgroup to Chimp.  There are some differences in sample groups, so also some differences are expected, but this result confirms what I got in the first test.

                                                             D stat         Z
Lithuanian     West-Finnish :  Norwegian      Chimp      0.0032     2.818
Belarusian      West-Finnish :  Norwegian      Chimp     -0.0026    -2.291
Bulgarian        West-Finnish :  Norwegian      Chimp     -0.0091    -7.949
Chuvash        West-Finnish  :  Norwegian      Chimp     -0.0248   -20.419

Aineiston lisäys sisältäen vepsät, karjalaiset ym.

Estonian Biocentre julkaisi uutta dataa, josta julkaisen nyt vain PCA-plotit.  Palaan asiaan jatkoanalyysien muodossa lähiaikoina, parin viikon kuluessa.

Lisäys 24.2.15

Testausta IMS-taustaisten keskinäisistä suhteista.  Lataa paketti tästä.  Dimensioiden 2 ja 3 Eigenvalue-arvot ovat kaikissa kuvissa hyvin lähellä toisiaan, joten dim. 2 ja 3 ovat täysin vertailukelpoisia, ja joskus asetelmat vaihtavat paikkaa.

A new study reveals more ancient East European migrations

This is really big news because we probably can in nearest future understand much better the origin of present-day Europeans, thanks for new ancient East European samples.   It is also sure that speculative flames will now go high,  but I am not going to take a stance on this before I have the new data, especially regarding new Yamna culture samples.  The reason to wait is that Haak et al. obviously uses same small and unclassified Finnish sample set than Lazaridis 2014 et al.  Anyway, here is the link, Haak et al.

http://biorxiv.org/content/biorxiv/early/2015/02/10/013433.full.pdf


I am also running a big European wide analysis using Chromopainter and Finestucture in linked mode.  It will be ready after two weeks or so.  Link to the software: http://www.maths.bris.ac.uk/~madjl/finestructure-old/chromopainter.html