Y-STR based Nevgen predictions of Polish Y-DNA hgs/SNPs labeled as K-M9

[Tomenable]

In this study (n=201), as many as 19% of all samples - 38 - were labeled as K-M9:

http://www.amsik.pl/archiwum/3_2013/3_13d.pdf

So far I have Nevgen-predicted 18 out of these 38. Here are the most likely results:

ID01 - N1c
ID33 - I2a1b2-L621
ID36 - I2a1b2-L621
ID43 - I2a1b2-L621
ID45 - I2a1b2-L621
ID51 - I1a3a2-BY351
ID60 - R1a1a1b2a-Z94
ID63 - R1a1a1b1a3a-L448
ID66 - R1a1a1b1a1-M458
ID67 - R1a1a1b1a2-Z280
ID74a - R1a1a1b1a2-Z280
ID74b - R1a1a1b1a2-Z280
ID81 - R1a1a1b1a1-M458
ID83 - R1a1a1b1a-YP694
ID87 - R1b1a1a2a1a1c2b1-L47
ID88 - R1b1a1a2a1a2e-DF19
ID89 - R1a1a1b1a1-M458
ID98 - R1a1a1b1a2-Z280

Twenty more K-M9 samples to go:

http://www.nevgen.org

ID106
ID111
ID112
ID113
ID129
ID143
ID144
ID150
ID151
ID157
ID158
ID173
ID179
ID180
ID181
ID184
ID185
ID186
ID187
ID188

 

[Tomenable]

And the remaining predictions are:

ID106 - R1b-L21
ID111 - R1b-PF7558
ID112 - R1a-Z92
ID113 - R1a-Z280
ID129 - R1a-M458
ID143 - R1a-YP694
ID144 - R1b-P312
ID150 - R1a-Y2613
ID151 - R1a-YP343
ID157 - R1a-YP4141
ID158 - I2a1b3-L621
ID173 - R1a-Y35
ID179 - R1b-DF99
ID180 - R1a-L1280
ID181 - R1a-L1280
ID184 - R1b-L47
ID185 - R1a-Z94 (probability of Z94 is 99,78% here)
ID186 - N1c
ID187 - R1a-Z284
ID188 - R1a-CTS1211

And one sample - ID109 - was classified as "unknown haplogroup" in the original study. Nevgen predicts it as:

ID109 - R1a-M458

 

[Tomenable]

In the original study, they had such results (n=201):

R1a = 119 (59,2%)
K-M9 = 38 (18,9%)
R1b = 30 (14,9%)
IJ = 13 (6,5%)
N/A = 1 (0,5%)

After predicting K-M9 & N/A with Nevgen (n=201):

R1a = 119 + 24 = 143 (71,1%)
R1b = 30 + 7 = 37 (18,4%)
IJ = 13 + 6 = 19 (9,5%)
N1c = 0 + 2 = 2 (1%)

Over 70 percent of R1a, more than in Sorbs and Kashubians!

 

[LeBrok]

If I understand well, these are contemporary residents of Greater Poland, right?

 

[Tomenable]

There is no difference betwen old residents and contemporary residents in Wielkopolska.

Greater Poland was not significantly affected by massive migrations in the 20th century. You are thinking about places such as Wrocław, where Polonized East Slavic origin of contemporary residents is evident (I have samples from Wrocław, and I can see elevated frequencies of I2a and E1b1b1; while lower frequency of R1a; actually before 1945 there was a lot of R1a in Lower Silesia*).

When we look at R1a frequencies, then what we can see is a west-east cline in Slavic areas.

That is, groups such as Kashubians, Greater Poles, Lusatian Sorbs have over 60% of R1a:

Kashubians - 63,4% (170 out of 268) ----> mostly Z280
Wielkopolans - 71,1% (143 out of 201) ----> equal Z280/M458
Lusatian Sorbs - 65,0% (80 out of 123) ----> mostly M458

When you go farther east, frequencies of R1a actually decline compared to those groups.

*I have pre-1945 samples from Lower Silesia, West Pomerania and East Prussia collected from FTDNA projects.

 

[LeBrok]

Kashubians - 63,4% (170 out of 268) ----> mostly Z280
Wielkopolans - 71,1% (143 out of 201) ----> equal Z280/M458
Lusatian Sorbs - 65,0% (80 out of 123) ----> mostly M458

What does it mean in Slavic terms?

 

[Tomenable]

What does it mean in Slavic terms?

We need to wait for more ancient DNA I guess.

 

[Tomenable]

Oh, it seems that there are huge problems with this sample set.

Michał from Anthrogenica wrote there is only 54,2% of R1a in this sample of 201 from Wielkopolska:

I'm waiting for Michał's next post to see why he thinks (if he really thinks so) that ~70% for Wielkopolska is wrong.

Yes, I am very strongly convinced that this number is wrong.
Firstly, let me note that there is absolutely no correlation between the SNP data provided by the authors and the STR-based haplogroup prediction for particular haplotypes (as performed by myself; I used the Whit Athey's predictor in all those cases when such prediction was beyond my competence). Secondly, the STR results for DYS437 are apparently wrong for a large group of 31 samples (ID125-ID155), and since this was probably an error made when constructing the table, we cannot be sure that the remaining STR data are correct. However, when ignoring those DYS437 results, the frequencies for predicted haplogroups are as follows:

R1a - 109/201 (54.2%)
including L260 - 41/201 (20.4%)
R1b - 39/201 (19.4%)
I1 - 21/201 (10.4%)
I2 - 20/201 (10.0%)
E1b - 3/201 (1.5%)
G2a - 2/201 (1%)
J1 - 2/201 (1%)
N - 2/201 (1%)
Q - 1/201 (0.5%)
L - 1/201 (0.5%)
T - 1/201 (0.5%)

As you can see, the overall frequency of R1a is more or less what one would expect for the entire country. However, the frequencies of L260, R1b, I1 and I2 seem to be a bit higher than the average values for the entire country (though I doubt such differences are statistically significant when using samples of this size).

 

[lyakh]

ID157 - R1a-YP4141 -interesting. So it appears to be M459- M198- M417- (very archaic sample of R1a)

 

[lyakh]

There are also four results which appear to be from R1a-Y2395 branch (YP694 is a descendant of Y2395):
ID63 - R1a1a1b1a3a-L448
ID83 - R1a1a1b1a-YP694
ID143 - R1a-YP694
ID187 - R1a-Z284

There are also two samples which appear to be R1a-Z94 - ID60 and ID185 (second has very high probability of being Z94+).

ID66, ID81, ID89, ID109, ID129 appear to be M458. ID66 appear to be L260 (has DYS395=10-14) and even YP256 (has DYS439=10). The same with ID81. But ID89 is probably L260- (has DYS385=11-14 and DYS439=11) - appears to be CTS11962+. ID109 also appears to have YP256. ID129 is probably L260 due to DYS385=10-14, but it may be not YP256+ (it has DYS439=11, YP256+ has rather DYS439=10).

 

[lyakh]

So L260 is probably in 45 of 201 samples which gives frequency of about 22,39%.
Not as much as in the case of Sorbs in Rębała's sample (about 35% of all appear to be L260+), but few times larger than among Kashubians in Rębała's (where only about 4% of all haplotypes appear to be L260+) and also significantly larger than in Kociewie region (where about 7% of samples appear to be L260+).