PDA

View Full Version : A first look at Kumtepe 6



Kurd
01-13-2016, 04:23 AM
After some manipulation, I finally have a somewhat workable sample for Kumtepe 6. It is a low coverage pseudo diploid, with low SNP overlap with Human Origins. The IBS run shows that Iceman is in the long lead with regards to similarity. Finally, it seems that Iceman has paired up with someone. This may mean that an Iceman/Kumtepe 6 component may be feasible in ADMIXTURE. I will do a couple of runs by the weekend to look into it. If a decent Kumtepe component does form, it would be interesting to see which W Asian/SE Europeans score high in it vs. Anatolian Farmers/ LBK. Formal analysis should follow.



NO
Sample / Pop
IBS Kumtepe 6


1
Kumtepe 6
100.00%


2
Iceman
70.24%


3
LBK_EN
68.55%


4
SATSURBILA
68.22%


5
RISE_baHu
68.19%


6
Stuttgart
68.09%


7
.DMXX
68.05%


8
Anatolia_Neolithic
67.88%


9
.Passa
67.78%


10
.Znertu
67.75%


11
Armenian
67.55%


12
RISE_baSin
67.50%


13
LaBrana1
67.49%


14
Greek
67.49%


15
KOTIAS
67.45%


16
.Anabasis
67.42%


17
Sicilian
67.42%


18
Bulgarian
67.37%


19
.Hanna
67.35%


20
.Zephyrous
67.34%


21
BedouinB
67.32%


22
Georgian
67.29%


23
Druze
67.25%


24
.Kurd_C2
67.24%


25
.Kurd_C3
67.23%


26
RISE_baAfan
67.22%


27
.NK19191
67.22%


28
.Zara
67.20%


29
.Kurd_C1
67.16%


30
Turkish
67.13%


31
Syrian
67.12%


32
Lithuanian
67.11%


33
Kurd_N
67.06%


34
Yemenite_Jew
67.00%


35
Srubnaya
66.99%


36
Russian
66.97%


37
Iranian
66.91%


38
Tajik_Pomiri
66.85%


39
Yemen
66.82%


40
.McNinja
66.81%


41
.TagHeuer
66.80%


42
.Rukha
66.79%


43
Yamnaya
66.78%


44
RISE_baAndrov
66.76%


45
.Sapporo
66.74%


46
.Kenji
66.73%


47
Karelia_HG
66.71%


48
Balochi
66.68%


49
Brahui
66.67%


50
.Sein
66.64%


51
.Pak_Gujjar
66.64%


52
Kostenki14
66.62%


53
Loschbour
66.61%


54
.Jesus
66.59%


55
Scythian_IA
66.58%


56
Makrani
66.56%


57
Pathan
66.56%


58
.Balq
66.55%


59
.Melif
66.52%


60
.Jam
66.51%


61
.Arbogan
66.48%


62
.Hanif
66.47%


63
.Bol_Nat
66.47%


64
.Jatt_Har393
66.45%


65
.Kandhari
66.41%


66
Pashtun_Afghan
66.40%


67
Sindhi
66.38%


68
.Kurd_SE
66.36%


69
Uzbek_Afghan
66.34%


70
Kalash
66.33%


71
.Farid
66.33%


72
.McNinjaMom
66.32%


73
.Khana
66.32%


74
.Varun
66.28%


75
Tajik_Afghan
66.27%


76
Uzbek
66.27%


77
.Dluffy
66.26%


78
.Gill
66.26%


79
RISE_baMezh
66.24%

Kurd
01-13-2016, 04:24 AM
80
.Bored
66.22%


81
.Kaido
66.10%


82
.Sadia
66.02%


83
.Parasar
65.86%


84
Bengali
65.83%


85
Hazara
65.75%


86
RISE_baKarasuk
65.64%


87
Puliyar
65.42%


88
Clovis
65.40%


89
Kusunda
65.31%


90
.Reza
65.27%


91
Ulchi
65.08%


92
Naga
65.03%


93
.Awale
64.96%


94
Mongola
64.94%


95
Han
64.76%


96
Dai
64.72%


97
Ami
64.72%


98
Atayal
64.68%


99
MA1
64.64%


100
Papuan
64.64%


101
Ust_Ishim
64.11%


102
.MOTA
63.21%


103
Yoruba
63.00%


104
Mbuti
62.48%


105
Denisovan
62.40%


106
Ancient_Denisova
62.18%

Anabasis
01-13-2016, 04:35 AM
kurd could you post any admixture result of it?

ZephyrousMandaru
01-13-2016, 04:52 AM
Interesting, will Kumtepe6 be used a reference to a future ADMIXTURE run and calculator or perhaps updated as a reference to an existing calculator?

vettor
01-13-2016, 06:11 AM
hmmm......I still do not even qualify on the charts fro this Anatolian!

Padre Organtino
01-13-2016, 10:04 AM
This guy doesn't look like your typical ENF or EEF.

Piquerobi
01-13-2016, 12:25 PM
Is it already available on Gedmatch?

Kurd
01-13-2016, 12:36 PM
This one shows the CHG shift of Kumtepe 6 and Iceman to the exclusion of neolithic farmers, such as Anatolians, LBK, and Stuttgart. Keep in mind that Kumtepe 6 is by far the least SNP overlapping sample in the study, and it is haploid. I am very confident that if that were not the case, Kumtepe 6 would be quite a bit higher up the list than Iceman.



NO
Sample / Pop
IBS SATSURB



1
SATSURBILA
100.00%


2
KOTIAS
70.22%


3
Iceman
68.63%


4
Kumtepe 6
68.22%


5
.Znertu
67.53%


6
Georgian
67.51%


7
RISE_baAfan
67.29%


8
.TagHeuer
67.27%


9
Armenian
67.26%


10
Kurd_N
67.22%


11
Yamnaya
67.22%


12
.Kurd_C1
67.16%


13
.NK19191
67.12%


14
.Arbogan
67.12%


15
.Anabasis
67.11%


16
Scythian_IA
67.10%


17
Kalash
67.07%


18
.Kurd_C2
67.07%


19
RISE_baSin
67.07%


20
.Sapporo
67.06%


21
.Bol_Nat
67.06%


22
.Kurd_C3
67.05%


23
Turkish
67.02%


24
.Hanna
67.00%


25
.Sein
67.00%


26
.Farid
66.99%


27
.DMXX
66.99%


28
.Jesus
66.98%


29
.Kurd_SE
66.98%


30
Druze
66.98%


31
Tajik_Pomiri
66.97%


32
.Melif
66.95%


33
Iranian
66.93%


34
RISE_baAndrov
66.92%


35
Balochi
66.91%


36
.Hanif
66.91%


37
Makrani
66.91%


38
Pathan
66.91%


39
Brahui
66.91%


40
Lithuanian
66.90%


41
.Khana
66.89%


42
Sindhi
66.88%


43
.Zephyrous
66.87%


44
Greek
66.87%


45
RISE_baMezh
66.86%


46
.Pak_Gujjar
66.86%


47
.Jam
66.85%


48
.Kandhari
66.84%


49
Bulgarian
66.82%


50
Sicilian
66.79%


51
Uzbek_Afghan
66.79%


52
Russian
66.78%


53
Syrian
66.78%


54
.Passa
66.77%


55
.Zara
66.76%


56
.McNinja
66.76%


57
Srubnaya
66.75%


58
Pashtun_Afghan
66.73%


59
LBK_EN
66.72%


60
Tajik_Afghan
66.71%


61
Stuttgart
66.71%


62
.Dluffy
66.69%


63
RISE_baHu
66.67%


64
.Rukha
66.66%


65
.Kaido
66.64%


66
.McNinjaMom
66.64%


67
.Balq
66.64%


68
.Gill
66.63%


69
.Kenji
66.62%


70
Yemenite_Jew
66.62%


71
.Bored
66.61%


72
.Jatt_Har393
66.60%


73
Anatolia_Neolithic
66.56%


74
BedouinB
66.54%


75
.Sadia
66.49%


76
Loschbour
66.44%


77
Uzbek
66.39%


78
Bengali
66.36%


79
Yemen
66.36%


80
Kostenki14
66.34%


81
.Parasar
66.34%


82
Karelia_HG
66.33%


83
RISE_baKarasuk
66.33%


84
.Varun
66.30%


85
LaBrana1
66.30%

Kurd
01-13-2016, 12:37 PM
86
Hazara
66.23%


87
.Reza
66.16%


88
Clovis
66.15%


89
Puliyar
65.88%


90
MA1
65.71%


91
Kusunda
65.60%


92
Mongola
65.50%


93
Papuan
65.50%


94
Ulchi
65.44%


95
Dai
65.40%


96
Naga
65.37%


97
Atayal
65.37%


98
Ami
65.35%


99
Han
65.27%


100
Ust_Ishim
65.24%


101
.Awale
64.71%


102
.MOTA
63.64%


103
Ancient_Denisova
63.19%


104
Mbuti
63.14%


105
Yoruba
63.11%


106
Denisovan
63.09%

Kurd
01-13-2016, 12:37 PM
Interesting, will Kumtepe6 be used a reference to a future ADMIXTURE run and calculator or perhaps updated as a reference to an existing calculator?

Yes, working on that now

Kurd
01-13-2016, 12:38 PM
This guy doesn't look like your typical ENF or EEF.

Indeed, he appears much more CHG shifted

Kurd
01-13-2016, 12:40 PM
Is it already available on Gedmatch?

No, I will not be able to upload it to Gedmatch. His file size is much below the Gedmatch requirement, due to small number of markers

Padre Organtino
01-13-2016, 02:13 PM
Indeed, he appears much more CHG shifted

I think we have a confirmation of secondary Neolithic expansion into Europe by Danube route that has lead to the formation of LBK culture. It fits nicely we the observed abundance of CHG in Central Europe compared to the scenario of Yamnaya-only source.

royking
01-13-2016, 02:18 PM
This is great! So Kurd has answered the question that his split commentaries on this forum--Kum6 indeed has a CHG component not reflected in the earliest Neolithic expansion into Europe.

Kurd
01-13-2016, 02:25 PM
I got the 1st results from a K11 Admixture run, with a Kumtepe6/Iceman component. The results were a little surprising in that there seemed to be a distinct pattern of selective W/SC/S Asian groups having a relatively high Kumtepe / ENF ratio. I will post in a couple of hours when I get a chance

Padre Organtino
01-13-2016, 02:39 PM
I got the 1st results from a K11 Admixture run, with a Kumtepe6/Iceman component. The results were a little surprising in that there seemed to be a distinct pattern of selective W/SC/S Asian groups having a relatively high Kumtepe / ENF ratio. I will post in a couple of hours when I get a chance

You mean like really different from CHG/ENF ratio distribution?

parasar
01-13-2016, 02:47 PM
This is great! So Kurd has answered the question that his split commentaries on this forum--Kum6 indeed has a CHG component not reflected in the earliest Neolithic expansion into Europe.

So you do not consider the Iceman/Otzi to be part of the earliest expansion into central Europe?

royking
01-13-2016, 02:56 PM
So you do not consider the Iceman/Otzi to be part of the earliest expansion into central Europe?
Yes, of course Otzi reflects the earliest expansion into Europe. The question is whether Otzi has an additional influence from later Neolithic expansions into Europe, like that influencing Kum6. The other possibility is that Impressed Ware in North Italy had a slightly different source population from Cardial Ware Iberia.

Anabasis
01-13-2016, 03:49 PM
Indeed, he appears much more CHG shifted

But but "dark green component does not appear in Yamnaya" :))

parasar
01-13-2016, 03:58 PM
Yes, of course Otzi reflects the earliest expansion into Europe. The question is whether Otzi has an additional influence from later Neolithic expansions into Europe, like that influencing Kum6. The other possibility is that Impressed Ware in North Italy had a slightly different source population from Cardial Ware Iberia.

With the caveat that Kumtepe6 is low coverage, the Iceman much to the west of Kum6 is showing more affinity to Satsurblia.

SATSURBILA 100.00%
KOTIAS 70.22%
Iceman 68.63%
Kumtepe 6 68.22%

royking
01-13-2016, 03:58 PM
But but "dark green component does not appear in Yamnaya" :))
I think the best explanation comes from a close analogy from psychology: CHG, personality constructs, like extraversion are aggregates--extraversion is a mixture of sociability, sensation-seeking, dominance etc...Extraversion is purely a mathematical construct summed from its internal components, but has validity as a predictor of other phenomena, like brain neurochemistry. CHG is also a mathematical/computational construct, exemplified by Kotias. However, like extraversion which in some studies correlates with dominance behavior, other studies, sociability--yet sociability may be somewhat independent from dominance and indeed have independent experimental correlates. So Yamnaya may associate with a different CHG subcomponent than Kum6's association with CHG.

Anabasis
01-13-2016, 04:08 PM
I think the best explanation comes from a close analogy from psychology: CHG, personality constructs, like extraversion are aggregates--extraversion is a mixture of sociability, sensation-seeking, dominance etc...Extraversion is purely a mathematical construct summed from its internal components, but has validity as a predictor of other phenomena, like brain neurochemistry. CHG is also a mathematical/computational construct, exemplified by Kotias. However, like extraversion which in some studies correlates with dominance behavior, other studies, sociability--yet sociability may be somewhat independent from dominance and indeed have independent experimental correlates. So Yamnaya may associate with a different CHG subcomponent than Kum6's association with CHG.


I know that. I agree with you. I wrote what you quote from me due to discussion there---> http://www.anthrogenica.com/showthread.php?6131-aDNA-from-Kumptepe-%28LATE-Neolithic-Anatolia%29&p=130931&viewfull=1#post130931

If you follow the discussion you will see the irony in my post.

Kurd
01-13-2016, 05:06 PM
Here is the FST matrix. Kumtepe appears equidistant to CHG, ENF, SW Asian, and WHG. EHG is not much behind. Don't pay much attention to the Denisovan component, as it has a significant margin of error, due to the SNPs being ascertained in Eurasians, and not in Denisovan and Africans. Nucleotides that are polymorphic in Eurasians may be monomorphic or ancestral in Africans and Denisovans, not to mention sequencing errors due to deamination and contamination issues with the Denisovan samples.




ENF
Denisovan
SE Asian
SW Asian
Kumtepe
EHG
WHG
Siberian
ASI
CHG


Denisovan
0.55











SE Asian
0.17
0.57










SW Asian
0.08
0.53
0.15









Kumtepe
0.15
0.61
0.22
0.14








EHG
0.10
0.54
0.14
0.09
0.16







WHG
0.10
0.54
0.15
0.08
0.14
0.08






Siberian
0.17
0.57
0.11
0.15
0.21
0.13
0.14





ASI
0.20
0.61
0.19
0.18
0.25
0.18
0.19
0.19




CHG
0.09
0.54
0.14
0.08
0.15
0.07
0.07
0.14
0.21



SSA
0.17
0.47
0.19
0.14
0.22
0.16
0.17
0.19
0.22
0.15

Kurd
01-13-2016, 05:15 PM
These are the K11 run results, focusing on the Kumtepe, ENF, and CHG components. This is sorted by Kumtepe, and the individuals have not been averaged into populations. The pattern will become clearer in the next table I post, which will have populations averaged. Nonetheless, the following can be observed:

1- My 4 Karasuk samples scored the highest from ancients. Karasuk scored 0 CHG and ENF (ENF is based on Anatolians and LBK, and CHG on Satsurbila, SC Asians, Kotias, and Caucausus groups)
2- From the moderns, Kurds score highest, as if Kumtepe was most ancestral to Kurds from all the W Asians. Some Brahui and Pathan samples come next, followed by a couple of S European samples. Surprisingly, Caucausus, Turks, Iranians, and some other W Asians are not on top where I had expected to see them



IND
Kumtepe
ENF
CHG


Kumtepe 6
99.99%
0.00%
0.00%


Iceman
99.99%
0.00%
0.00%


RISE_baKarasuk
70.57%
0.00%
0.00%


RISE_baKarasuk
32.62%
0.00%
0.00%


Srubnaya
32.08%
0.00%
0.00%


RISE_baSin
31.26%
0.00%
0.77%


RISE_baKarasuk
22.73%
0.00%
0.00%


RISE_baKarasuk
21.46%
0.00%
13.17%


LBK_EN
19.21%
80.54%
0.00%


LBK_EN
17.76%
81.35%
0.00%


RISE_irRus
15.72%
0.00%
5.98%


.Kurd_C3
15.37%
19.12%
46.73%


Srubnaya
14.13%
8.96%
2.74%


RISE_baAndrov
13.22%
0.00%
1.18%


Yamnaya
12.48%
0.00%
21.58%


.Kurd_C2
12.01%
18.23%
43.99%


RISE_baHu
11.65%
11.13%
0.00%


.Kurd_C1
11.15%
19.76%
42.45%


Brahui
10.92%
1.63%
50.80%


Druze
10.50%
18.57%
26.56%


French
10.44%
29.65%
7.95%


.Kurd_SE1
10.33%
12.79%
51.12%


Cypriot
10.18%
28.10%
17.28%


Pathan
10.12%
8.33%
45.79%


Pathan
9.92%
0.00%
52.72%


LBK_EN
9.83%
88.00%
0.00%


Turkmen_Afghan
9.79%
7.60%
29.18%


.Farid
9.50%
2.55%
42.66%


Greek
9.48%
33.42%
15.68%


BedouinA
9.41%
23.24%
19.80%


Kumyk
9.39%
15.71%
34.81%


Iranian
9.38%
13.69%
31.82%


Russian
9.24%
17.34%
4.07%


Pathan
9.24%
2.74%
47.77%


Balochi
9.22%
3.84%
46.70%

Padre Organtino
01-13-2016, 05:15 PM
Kumtepe is both less Basal and more MA shifted than ENF and CHG judging from this.

Not sure what we're seeing here.

Padre Organtino
01-13-2016, 05:19 PM
Really interested in score for Caucasus now.

Kurd
01-13-2016, 05:24 PM
The pattern becomes a little clearer when the populations are averaged. The couple of Pathan and Brahui samples that scored high Kumtepe individually, have been consumed by their population averages further down the list. Next to get posted are ratios



IND
Kumtepe
ENF
CHG


Iceman
99.99%
0.00%
0.00%


Kumtepe 6
99.99%
0.00%
0.00%


RISE_baSin
31.26%
0.00%
0.77%


RISE_baKarasuk
25.71%
0.00%
4.09%


Srubnaya
16.50%
6.68%
4.36%


RISE_irRus
15.72%
0.00%
5.98%


.Kurd_C3
15.37%
19.12%
46.73%


.Kurd_C2
12.01%
18.23%
43.99%


LBK_EN
11.70%
87.47%
0.00%


RISE_baHu
11.65%
11.13%
0.00%


.Kurd_C1
11.15%
19.76%
42.45%


.Kurd_SE1
10.33%
12.79%
51.12%


.Farid
9.50%
2.55%
42.66%


RISE_baAndrov
8.75%
0.85%
2.93%


.Znertu
8.34%
19.86%
30.17%


KOTIAS
8.31%
2.11%
60.92%


.Pak_Gujjar
8.03%
3.39%
48.00%


.Melif
7.55%
8.42%
44.95%


.Kurd_SE2
7.54%
4.78%
53.49%


.Kaido
7.49%
6.28%
44.15%


Yamnaya
6.83%
0.00%
30.33%


.Zara
6.75%
3.14%
49.66%


.Hanna
6.52%
22.24%
31.98%


.Zephyrous
6.42%
25.88%
37.57%


.DMXX
6.32%
18.36%
35.25%


RISE_baMezh
6.31%
0.00%
8.93%


.Passa
6.15%
40.53%
10.13%


.Sein
6.02%
1.95%
46.45%


Kumyk
5.86%
17.65%
34.45%


.McNinja
5.31%
0.00%
48.49%


Maltese
5.26%
31.71%
13.40%


Chechen
5.26%
16.47%
31.78%


Iranian_Jew
5.23%
23.68%
33.51%


Tajik_Pomiri
5.19%
8.52%
37.23%


.Khana
5.09%
2.79%
61.00%


.Gill
5.05%
5.12%
49.83%


Cypriot
4.93%
30.38%
21.54%

Padre Organtino
01-13-2016, 05:24 PM
Hm, hm.

If we are to believe the BA_Karasyuk score we have a specimen that was related to the crazy Dene-Caucasian language macro family))

Kurd
01-13-2016, 05:25 PM
English
4.87%
27.61%
7.43%


.Kandhari
4.86%
6.66%
41.49%


Syrian
4.62%
21.30%
24.03%


.McNinjaMom
4.58%
0.55%
54.14%


.Hanif
4.56%
4.80%
57.60%


Bulgarian
4.52%
27.74%
11.54%


Kurd_N
4.48%
18.59%
35.05%


Turkmen_Afghan
4.44%
8.85%
23.22%


Georgian_Jew
4.34%
23.03%
32.50%


.Arbogan
4.30%
20.39%
35.96%


Greek
4.30%
30.92%
15.54%


Tuscan
4.24%
33.42%
11.87%


.Balq
4.13%
3.59%
42.88%


Pathan
4.02%
6.04%
47.54%


Georgian
3.88%
22.56%
39.44%


.Jatt_Har393
3.85%
3.18%
46.45%


Yemen
3.83%
15.48%
16.72%


Balochi
3.82%
5.76%
54.44%


Icelandic
3.80%
25.59%
5.91%


French
3.79%
29.13%
7.16%


Poltavka
3.76%
0.00%
44.17%


Stuttgart
3.70%
59.90%
0.00%


Druze
3.69%
25.11%
26.41%


Norwegian
3.64%
23.26%
5.12%


Sicilian
3.63%
32.93%
14.54%


Tajik_Afghan
3.60%
8.45%
33.67%


.Varun
3.60%
0.34%
41.72%


BedouinA
3.53%
20.34%
16.41%


.NK19191
3.49%
9.07%
37.94%

Kurd
01-13-2016, 05:26 PM
Iraqi_Jew
3.46%
24.68%
29.51%


Spanish
3.41%
34.15%
4.70%


Brahui
3.36%
5.44%
56.41%


Uzbek
3.36%
4.95%
22.36%


Lezgin
3.30%
14.80%
36.32%


Abkhasian
3.27%
20.98%
39.99%


Jordanian
3.26%
24.27%
21.22%


Scottish
3.17%
25.14%
4.95%


Grand Total
3.17%
16.19%
18.85%


Turkmen
3.16%
11.93%
25.44%


Croatian
3.13%
27.65%
8.77%


Adygei
3.13%
18.83%
34.41%


Sardinian
3.10%
49.40%
0.45%


Estonian
3.02%
18.85%
4.01%


Turkish
3.01%
22.62%
24.73%


Iranian
3.01%
15.75%
36.51%


Makrani
2.97%
5.23%
51.66%


Belarusian
2.96%
21.71%
9.02%


Hazara_Afghan
2.94%
3.83%
30.78%


.Sapporo
2.91%
0.00%
56.40%


Finnish
2.83%
17.29%
2.46%


Albanian
2.81%
33.51%
11.58%


Sindhi
2.81%
3.75%
51.39%


Saudi
2.74%
15.41%
9.62%


Armenian
2.68%
25.85%
31.88%


Pashtun_Afghan
2.63%
6.24%
42.53%


Czech
2.60%
23.89%
8.95%


Hungarian
2.57%
25.90%
8.07%


Lebanese
2.57%
26.65%
21.55%


Yemenite_Jew
2.55%
22.61%
15.96%


.Bol_Nat
2.51%
2.47%
44.58%


.Awale
2.49%
4.56%
0.00%


Burusho
2.46%
3.38%
43.89%


French_South
2.33%
36.01%
4.39%


.Bored
2.30%
0.46%
44.47%


Lithuanian
2.24%
19.75%
4.20%


.Jesus
2.18%
21.04%
35.92%


Uygur
2.14%
4.60%
20.58%


Scythian_IA
2.12%
3.18%
28.35%


Clovis
2.12%
0.00%
0.00%


Russian
2.08%
17.97%
5.91%


.Kenji
1.96%
1.66%
47.55%


.Parasar
1.87%
3.07%
47.95%


Hazara
1.77%
4.20%
20.33%


Altaian
1.77%
3.50%
7.98%


.Rukha
1.75%
11.30%
43.25%


Kyrgyz
1.75%
5.67%
9.01%


Ukrainian
1.73%
21.97%
8.05%


.Jam
1.57%
5.83%
52.94%


Sherpa
1.23%
0.40%
11.35%


.Dluffy
1.20%
0.75%
49.20%


Great_Andamanese
1.13%
0.06%
12.79%


Bengali
1.08%
1.05%
42.18%


Masai
0.89%
1.62%
1.08%


Ust_Ishim
0.71%
4.22%
16.40%


Uzbek_Afghan
0.70%
7.80%
37.87%


Tibet-refugees
0.60%
0.16%
2.08%


BedouinB
0.47%
1.34%
0.62%


Kalash
0.38%
0.24%
2.17%


Kusunda
0.32%
0.00%
12.20%


Mongola
0.30%
0.15%
0.74%


Ulchi
0.12%
0.00%
0.02%


Puliyar
0.11%
0.00%
33.63%


.Sadia
0.01%
0.00%
57.14%


Han
0.00%
0.00%
0.06%


.MOTA
0.00%
0.00%
6.71%


.Reza
0.00%
0.00%
48.25%


Ami
0.00%
0.00%
0.00%


Anatolia_Neolithic
0.00%
99.54%
0.00%


Ancient_Denisova
0.00%
0.00%
0.00%

Kurd
01-13-2016, 05:27 PM
Atayal
0.00%
0.00%
0.00%


Dai
0.00%
0.00%
0.03%


Denisovan
0.00%
0.00%
0.00%


Esan
0.00%
0.00%
0.00%


Ho
0.00%
0.00%
27.76%


Karelia_HG
0.00%
0.00%
0.00%


Kostenki14
0.00%
4.75%
0.40%


LaBrana1
0.00%
0.00%
0.00%


Loschbour
0.00%
0.00%
0.00%


MA1
0.00%
0.00%
0.00%


Naga
0.00%
0.00%
0.00%


Nganasan
0.00%
0.00%
0.00%


Onge
0.00%
0.00%
0.00%

Padre Organtino
01-13-2016, 05:35 PM
Actually no, he's more Basal than CHG and ENF, I was wrong.

Padre Organtino
01-13-2016, 05:35 PM
Can we fit him as a simple CHG and ENF mix?

Anabasis
01-13-2016, 05:48 PM
Kurd could you run your calculators in gedmatch results for kum6?

Kurd
01-13-2016, 05:55 PM
Can we fit him as a simple CHG and ENF mix?

We will have to see how it does when I formally analyze. CHG and ENF seem reasonable. Not sure if we need additional WHG type

Kurd
01-13-2016, 05:57 PM
Kurd could you run your calculators in gedmatch results for kum6?

Can't upload since he doesn't meet file size requirements for Gedmatch due to lower number of markers

Padre Organtino
01-13-2016, 05:58 PM
We will have to see how it does when I formally analyze. CHG and ENF seem reasonable. Not sure if we need additional WHG type

It's just that this component seems all over the place with a rather illogical (from the first glance) distribution.

ArmandoR1b
01-13-2016, 06:10 PM
So Yamnaya may associate with a different CHG subcomponent than Kum6's association with CHG.
If they are different they should have different names.

ArmandoR1b
01-13-2016, 06:16 PM
The pattern becomes a little clearer when the populations are averaged. The couple of Pathan and Brahui samples that scored high Kumtepe individually, have been consumed by their population averages further down the list. Next to get posted are ratios

This is very close to what I expected to see. Kumtepe and CHG are two different components and CHG is 0.00% in Kumtepe and KOTIAS and Yamnaya are very low in Kumtepe but very high in CHG. That means the component in Kumtepe needs to have a different name than CHG because they are two different components. This supports the argument I've had all along.

Anabasis
01-13-2016, 06:16 PM
It's just that this component seems all over the place with a rather illogical (from the first glance) distribution.

Well but you can run it in R rigt?

Kurd
01-13-2016, 06:38 PM
This is very close to what I expected to see. Kumtepe and CHG are two different components and CHG is 0.00% in Kumtepe and KOTIAS and Yamnaya are very low in Kumtepe but very high in CHG. That means the component in Kumtepe needs to have a different name than CHG because they are two different components. This supports the argument I've had all along.

Satsurbila can't be considered since he is declared CHG ancestral at 100%. Kotias scored 8%. Since Satsurbila predates Kumtepe, I would like to run Kumtepe through a CHG calculator where Kum is not declared ancestral. Also formal stats should shed more light

Anabasis
01-13-2016, 07:38 PM
This is very close to what I expected to see. Kumtepe and CHG are two different components and CHG is 0.00% in Kumtepe and KOTIAS and Yamnaya are very low in Kumtepe but very high in CHG. That means the component in Kumtepe needs to have a different name than CHG because they are two different components. This supports the argument I've had all along.

Man he isolated Kumtepe and create a component. Thats why while kum6 has %99his own component but Kotias or Satsurbila 0. What is interesting is Anatolian Neolethic is %0 kum too. How can you explain this then? If Kurd run kum6 sample in his calculators Euroasia K10 CHG , K11 CHG-NAF or K9 ASI we will understand whats going on with kumtepe.

ArmandoR1b
01-13-2016, 07:44 PM
Man he isolated Kumtepe and create a component. Thats why while kum6 has %99his own component but Kotias or Satsurbila 0. What is interesting is Anatolian Neolethic is %0 kum too. How can you explain this then? If Kurd run kum6 sample in his calculators Euroasia K10 CHG , K11 CHG-NAF or K9 ASI we will understand whats going on with kumtepe.

Kumtepe, Anatolian Neolithic, and CHG are three different components. That has been my point all along.

Kurd
01-13-2016, 09:35 PM
Here we have the ratio of Kumtepe / (Kumtepe + ENF), which gives us an idea of the relative amount of Kumtepe admixture with regards to the total of ENF + Kumtepe


IND / POP
Kumtepe / (Kumtepe+ENF)


Iceman
100.0%


Kumtepe 6
100.0%


RISE_baSin
100.0%


RISE_baKarasuk
100.0%


RISE_irRus
100.0%


Yamnaya
100.0%


RISE_baMezh
100.0%


.McNinja
100.0%


Poltavka
100.0%


.Sapporo
100.0%


Clovis
100.0%


Kusunda
99.7%


Ulchi
99.2%


Puliyar
99.1%


Great_Andamanese
95.1%


.Sadia
93.1%


.Varun
91.4%


RISE_baAndrov
91.2%


.McNinjaMom
89.2%


.Bored
83.3%


KOTIAS
79.7%


Han
79.5%


Tibet-refugees
79.2%


.Farid
78.8%


.Sein
75.6%


Sherpa
75.4%


Srubnaya
71.2%


.Pak_Gujjar
70.3%


.Zara
68.3%


Mongola
66.4%


.Khana
64.6%


.Dluffy
61.4%


.Kurd_SE2
61.2%


Kalash
60.7%


.Jatt_Har393
54.8%


.Kaido
54.4%


.Kenji
54.1%


.Balq
53.5%


RISE_baHu
51.1%


Bengali
50.5%


.Bol_Nat
50.4%


.Gill
49.6%


.Hanif
48.7%


.Melif
47.3%


.Kurd_SE1
44.7%


.Kurd_C3
44.6%


Hazara_Afghan
43.4%


Sindhi
42.9%


.Kandhari
42.2%


Burusho
42.1%


Uzbek
40.4%


Scythian_IA
40.0%


Pathan
40.0%


Balochi
39.9%


.Kurd_C2
39.7%


Brahui
38.2%


Tajik_Pomiri
37.9%


.Parasar
37.8%


Makrani
36.2%


.Kurd_C1
36.1%


Masai
35.4%


.Awale
35.3%


Altaian
33.6%


Turkmen_Afghan
33.4%


Uygur
31.8%


Tajik_Afghan
29.9%


Hazara
29.7%


Pashtun_Afghan
29.6%


.Znertu
29.6%


.NK19191
27.8%


BedouinB
26.1%


.DMXX
25.6%


Kumyk
24.9%


Chechen
24.2%


Kyrgyz
23.6%


.Hanna
22.7%


.Jam
21.2%


Turkmen
21.0%


.Zephyrous
19.9%


Yemen
19.8%


Kurd_N
19.4%


Lezgin
18.3%


Iranian_Jew
18.1%


Syrian
17.8%


.Arbogan
17.4%


Iranian
16.0%


Georgian_Jew
15.9%


Saudi
15.1%


English
15.0%


BedouinA
14.8%


Georgian
14.7%


Ust_Ishim
14.5%


Adygei
14.2%


Maltese
14.2%


Finnish
14.1%

Kurd
01-13-2016, 09:36 PM
Bulgarian
14.0%


Cypriot
14.0%


Estonian
13.8%


Norwegian
13.5%


Abkhasian
13.5%


.Rukha
13.4%


.Passa
13.2%


Icelandic
12.9%


Druze
12.8%


Iraqi_Jew
12.3%


Greek
12.2%


Belarusian
12.0%


Jordanian
11.9%


LBK_EN
11.8%


Turkish
11.8%


French
11.5%


Tuscan
11.3%


Scottish
11.2%


Russian
10.4%


Lithuanian
10.2%


Croatian
10.2%


Yemenite_Jew
10.1%


Sicilian
9.9%


Czech
9.8%


.Jesus
9.4%


Armenian
9.4%


Spanish
9.1%


Hungarian
9.0%


Lebanese
8.8%


Uzbek_Afghan
8.2%


Albanian
7.7%


Ukrainian
7.3%


French_South
6.1%


Sardinian
5.9%


Stuttgart
5.8%


Papuan
0.1%


Kostenki14
0.0%


Anatolia_Neolithic
0.0%


SATSURBILA
0.0%


.Reza
0.0%


Paniyas
0.0%


Ho
0.0%


RISE_baAfan
0.0%


.MOTA
0.0%


Dai
0.0%


Ami
0.0%


Ancient_Denisova
0.0%


Atayal
0.0%


Denisovan
0.0%


Esan
0.0%


Karelia_HG
0.0%


LaBrana1
0.0%


Loschbour
0.0%


MA1
0.0%


Naga
0.0%


Nganasan
0.0%


Onge
0.0%


Yoruba
0.0%

Kurd
01-13-2016, 09:48 PM
A few individual run results. Don't pay much attention to the Denisovan score for the reasons previously mentioned



IND
ENF
Denisovan
SE Asian
SW Asian
Kumtepe
EHG
WHG
Siberian
ASI
CHG
SSA


.Arbogan
20.39%
0.75%
0.00%
20.71%
4.30%
9.07%
3.15%
4.27%
1.40%
35.96%
0.00%


.Awale
4.56%
0.62%
0.00%
36.19%
2.49%
0.00%
1.67%
0.00%
1.41%
0.00%
53.06%


.Balq
3.59%
0.52%
6.22%
0.00%
4.13%
9.64%
13.28%
7.95%
11.79%
42.88%
0.00%


.Bol_Nat
2.47%
0.37%
4.99%
0.00%
2.51%
16.84%
9.58%
3.92%
14.75%
44.58%
0.00%


.Bored
0.46%
0.00%
11.25%
1.96%
2.30%
9.06%
10.76%
0.00%
19.74%
44.47%
0.00%


.Dluffy
0.75%
0.70%
5.98%
0.00%
1.20%
9.67%
13.95%
1.06%
17.49%
49.20%
0.00%


.DMXX
18.36%
0.00%
3.95%
19.33%
6.32%
3.75%
6.87%
1.55%
3.76%
35.25%
0.86%


.Farid
2.55%
0.28%
0.00%
18.79%
9.50%
8.53%
1.15%
0.00%
11.52%
42.66%
5.01%


.Gill
5.12%
0.00%
1.45%
1.83%
5.05%
7.35%
7.44%
3.43%
18.35%
49.83%
0.16%


.Hanif
4.80%
1.14%
7.51%
0.00%
4.56%
0.11%
6.10%
1.83%
16.35%
57.60%
0.00%


.Hanna
22.24%
0.37%
4.04%
15.95%
6.52%
0.00%
13.87%
5.02%
0.00%
31.98%
0.00%


.Jam
5.83%
0.43%
4.23%
0.86%
1.57%
10.59%
6.37%
0.00%
13.88%
52.94%
3.31%


.Jatt_Har393
3.18%
0.04%
4.14%
0.00%
3.85%
12.39%
11.74%
3.19%
15.03%
46.45%
0.00%


.Jesus
21.04%
0.00%
1.45%
14.91%
2.18%
8.31%
1.22%
3.66%
4.78%
35.92%
6.54%


.Kaido
6.28%
0.00%
4.97%
2.77%
7.49%
12.99%
7.00%
2.72%
11.64%
44.15%
0.00%


.Kandhari
6.66%
0.00%
6.24%
8.42%
4.86%
9.52%
10.43%
0.00%
9.95%
41.49%
2.43%


.Kenji
1.66%
1.30%
6.93%
1.42%
1.96%
14.55%
2.01%
3.99%
18.62%
47.55%
0.00%


.Khana
2.79%
0.00%
7.01%
0.00%
5.09%
0.00%
3.03%
2.90%
18.17%
61.00%
0.00%

Kurd
01-13-2016, 09:50 PM
IND
ENF
Denisovan
SE Asian
SW Asian
Kumtepe
EHG
WHG
Siberian
ASI
CHG
SSA


.Kurd_C1
19.76%
0.24%
0.00%
19.79%
11.15%
0.87%
1.92%
2.10%
1.34%
42.45%
0.39%


.Kurd_C2
18.23%
0.53%
3.21%
18.57%
12.01%
3.46%
0.00%
0.00%
0.00%
43.99%
0.00%


.Kurd_C3
19.12%
0.55%
0.12%
16.83%
15.37%
0.00%
0.00%
0.62%
0.65%
46.73%
0.00%


.Kurd_SE
4.78%
0.00%
2.08%
5.13%
7.54%
9.03%
0.00%
1.71%
12.88%
53.49%
3.36%


.McNinja
0.00%
0.68%
0.00%
1.82%
5.31%
8.43%
9.91%
3.24%
22.12%
48.49%
0.00%


.McNinjaMom
0.55%
0.00%
3.04%
0.00%
4.58%
4.49%
11.04%
2.41%
19.75%
54.14%
0.00%


.Melif
8.42%
0.00%
0.00%
3.70%
7.55%
11.58%
7.42%
7.21%
8.88%
44.95%
0.30%


.MOTA
0.00%
5.00%
0.00%
3.17%
0.00%
0.00%
0.00%
1.17%
6.42%
6.71%
77.53%


.NK19191
9.07%
0.00%
0.52%
26.62%
3.49%
7.27%
11.42%
1.34%
2.33%
37.94%
0.00%


.Pak_Gujjar
3.39%
0.73%
2.95%
0.00%
8.03%
13.96%
1.48%
0.00%
18.82%
48.00%
2.63%


.Parasar
3.07%
0.34%
3.60%
0.00%
1.87%
11.81%
6.78%
3.23%
19.46%
47.95%
1.90%


.Passa
40.53%
0.00%
0.00%
13.53%
6.15%
2.68%
26.07%
0.00%
0.41%
10.13%
0.49%


.Reza
0.00%
1.16%
13.05%
0.18%
0.00%
4.29%
1.40%
7.42%
23.38%
48.25%
0.85%


.Rukha
11.30%
0.00%
0.00%
4.92%
1.75%
15.24%
9.35%
4.22%
9.96%
43.25%
0.00%


.Sadia
0.00%
0.00%
1.94%
2.95%
0.01%
5.99%
3.40%
5.16%
20.51%
57.14%
2.91%


.Sapporo
0.00%
0.12%
0.00%
3.72%
2.91%
12.03%
1.94%
5.55%
16.93%
56.40%
0.40%


.Sein
1.95%
0.00%
3.76%
12.10%
6.02%
9.27%
4.24%
2.79%
13.42%
46.45%
0.00%


.Varun
0.34%
1.35%
3.54%
0.00%
3.60%
14.46%
7.48%
6.71%
20.81%
41.72%
0.00%


.Zara
3.14%
0.85%
2.39%
21.90%
6.75%
0.00%
4.26%
0.00%
8.47%
49.66%
2.57%


.Zephyrous
25.88%
0.52%
0.00%
26.37%
6.42%
0.00%
0.00%
0.00%
3.23%
37.57%
0.00%


.Znertu
19.86%
0.00%
0.83%
17.88%
8.34%
11.19%
7.97%
0.00%
3.23%
30.17%
0.53%

Dr_McNinja
01-13-2016, 10:51 PM
IND
ENF
Denisovan
SE Asian
SW Asian
Kumtepe
EHG
WHG
Siberian
ASI
CHG
SSA


.Kurd_C1
19.76%
0.24%
0.00%
19.79%
11.15%
0.87%
1.92%
2.10%
1.34%
42.45%
0.39%


.Kurd_C2
18.23%
0.53%
3.21%
18.57%
12.01%
3.46%
0.00%
0.00%
0.00%
43.99%
0.00%


.Kurd_C3
19.12%
0.55%
0.12%
16.83%
15.37%
0.00%
0.00%
0.62%
0.65%
46.73%
0.00%


.Kurd_SE
4.78%
0.00%
2.08%
5.13%
7.54%
9.03%
0.00%
1.71%
12.88%
53.49%
3.36%


.McNinja
0.00%
0.68%
0.00%
1.82%
5.31%
8.43%
9.91%
3.24%
22.12%
48.49%
0.00%


.McNinjaMom
0.55%
0.00%
3.04%
0.00%
4.58%
4.49%
11.04%
2.41%
19.75%
54.14%
0.00%


.Melif
8.42%
0.00%
0.00%
3.70%
7.55%
11.58%
7.42%
7.21%
8.88%
44.95%
0.30%


.MOTA
0.00%
5.00%
0.00%
3.17%
0.00%
0.00%
0.00%
1.17%
6.42%
6.71%
77.53%


.NK19191
9.07%
0.00%
0.52%
26.62%
3.49%
7.27%
11.42%
1.34%
2.33%
37.94%
0.00%


.Pak_Gujjar
3.39%
0.73%
2.95%
0.00%
8.03%
13.96%
1.48%
0.00%
18.82%
48.00%
2.63%


.Parasar
3.07%
0.34%
3.60%
0.00%
1.87%
11.81%
6.78%
3.23%
19.46%
47.95%
1.90%


.Passa
40.53%
0.00%
0.00%
13.53%
6.15%
2.68%
26.07%
0.00%
0.41%
10.13%
0.49%


.Reza
0.00%
1.16%
13.05%
0.18%
0.00%
4.29%
1.40%
7.42%
23.38%
48.25%
0.85%


.Rukha
11.30%
0.00%
0.00%
4.92%
1.75%
15.24%
9.35%
4.22%
9.96%
43.25%
0.00%


.Sadia
0.00%
0.00%
1.94%
2.95%
0.01%
5.99%
3.40%
5.16%
20.51%
57.14%
2.91%


.Sapporo
0.00%
0.12%
0.00%
3.72%
2.91%
12.03%
1.94%
5.55%
16.93%
56.40%
0.40%


.Sein
1.95%
0.00%
3.76%
12.10%
6.02%
9.27%
4.24%
2.79%
13.42%
46.45%
0.00%


.Varun
0.34%
1.35%
3.54%
0.00%
3.60%
14.46%
7.48%
6.71%
20.81%
41.72%
0.00%


.Zara
3.14%
0.85%
2.39%
21.90%
6.75%
0.00%
4.26%
0.00%
8.47%
49.66%
2.57%


.Zephyrous
25.88%
0.52%
0.00%
26.37%
6.42%
0.00%
0.00%
0.00%
3.23%
37.57%
0.00%


.Znertu
19.86%
0.00%
0.83%
17.88%
8.34%
11.19%
7.97%
0.00%
3.23%
30.17%
0.53%


How many markers did this utilize?

Gravetto-Danubian
01-13-2016, 11:29 PM
The pattern becomes a little clearer when the populations are averaged. The couple of Pathan and Brahui samples that scored high Kumtepe individually, have been consumed by their population averages further down the list. Next to get posted are ratios



Kurd
Well done on your prodigious work

I know you had been previously locked in debate (http://www.anthrogenica.com/showthread.php?5387-Interesting-Ancient-Dstats/page10) about the significance of Oetzi. Although im sure further testing and samples are warranted, your findings are very important.

I think it's early days to correlate this into a grand narrative, but for decades archaeologists had been positing a second "Neolithic wave" into Europe, and tidbits here and there (eg Romanian Neolithic aDNA paper by Harvella, and the abstract on Neolithic- Bronze Age Greece by Unterlander) strongly implied this on the basis of mtDNA. So we are now seeing it from a genome-wide perspective.

it would be nice to clarify where this second wave emanated from - but if the first one was the classical Levantine - Aegean route, then this second one might be a more overland one via through central Anatolia - NW Anatolia - Bulgaria (where more hunter-gatherers were encountered) - possibly emanating beyond the traditional Neolithic 'core areas'- ? Iran ? If, so, I wonder if there were two waves toward the east (ie India) also ?

Whatever the case, it is interesting that the CHG in Yamnaya appears 1/3 reduced - from ~ 33% to 21%.
Even more noteworthy, however, is that its successors (ie replacements) - Srubnaya - Andronovo - have now reduced to ~ 0 % CHG, replaced wholly by this newfound component.

To Europe - I see BA Hungary sample also has this component (11%, with 11% ENF, 0% Kotian CHG, and much of the remaining WHG and non-Yamnaya EHG). I'd imagine Mycenean (and immediate pre-Mycenean) Greece would look the same.

Quite clearly, we are looking at a significant exchange on the steppe, possibly by Indo-European groups coming from - wherever these Srubnaya - Andronovo- BA Souteastern European people came from.

When and if you can - can you add some Bell Beaker, CWC, Unetice and the BA _Montenegro sample (from Allentoft) to your K11 runs ?

Padre Organtino
01-14-2016, 08:32 AM
Whatever the case, it is interesting that the CHG in Yamnaya appears 1/3 reduced - from ~ 33% to 21%.
Even more noteworthy, however, is that its successors (ie replacements) - Srubnaya - Andronovo - have now reduced to ~ 0 % CHG, replaced wholly by this newfound component.

To Europe - I see BA Hungary sample also has this component (11%, with 11% ENF, 0% Kotian CHG, and much of the remaining WHG and non-Yamnaya EHG). I'd imagine Mycenean (and immediate pre-Mycenean) Greece would look the same.

Quite clearly, we are looking at a significant exchange on the steppe, possibly by Indo-European groups coming from - wherever these Srubnaya - Andronovo- BA Souteastern European people came from.

When and if you can - can you add some Bell Beaker, CWC, Unetice and the BA _Montenegro sample (from Allentoft) to your K11 runs ?

That's why it is very likely a phantom component. I don't see how they could be a very different mix from Yamnaya and contain no CHG

Gravetto-Danubian
01-14-2016, 09:08 AM
That's why it is very likely a phantom component. I don't see how they could be a very different mix from Yamnaya and contain no CHG

It is of course possible
That's why I proviso'd - if confirmed ;)
So I look forward to further discussions, although I have little reason to doubt it. We have seen population flux and migrations occurred everywhere. With its climactic instability, low population densities, competition for water and preferential access to goods from "higher centres" in the Balkans and beyond the Caucasus, the steppe is IMO the prime place for such turnovers.

Padre Organtino
01-14-2016, 10:30 AM
It is of course possible
That's why I proviso'd - if confirmed ;)
So I look forward to further discussions, although I have little reason to doubt it. We have seen population flux and migrations occurred everywhere. With its climactic instability, low population densities, competition for water and preferential access to goods from "higher centres" in the Balkans and beyond the Caucasus, the steppe is IMO the prime place for such turnovers.

For this to be true you need to have a pure unmixed EHG guy meeting some Kumtepe chicks in the steppe. I am not saying that this is impossible per se but so far we have only seen the trend towards EHG dilution by CHG over time. Plus one need a lot of imagination to come up with a scenario where this supposed population roams between Anatolia and Central Asia but leaves almost no noticeable impact upon locals with the exception of several ethnic groups.
Me thinks this is not a very plausible scenario.

Gravetto-Danubian
01-15-2016, 06:14 AM
For this to be true you need to have a pure unmixed EHG guy meeting some Kumtepe chicks in the steppe. I am not saying that this is impossible per se but so far we have only seen the trend towards EHG dilution by CHG over time. Plus one need a lot of imagination to come up with a scenario where this supposed population roams between Anatolia and Central Asia but leaves almost no noticeable impact upon locals with the exception of several ethnic groups.
Me thinks this is not a very plausible scenario.

For what to be true, exactly ? For the turnover in the steppe during the Yamnaya -> catacomb -> Srubnaya transition ? No qualifications are required - it happened.

Kurd
01-16-2016, 12:20 AM
Update

I have not posted anything in the past few days, as I scrapped all my databases and am starting from scratch, with a renewed focus on creating a couple of robust databases, with an emphasis on maximizing markers and overlap with the thousands of samples I have. I am hoping this will increase accuracy, and I can pick up where I left off with Kumtepe.

This is a rather laborious time intensive process, and I anticipate being up and ready in a few days.

In the pipeline is also the idea of branching out of Eurasian ascertained SNP arrays, and doing analysis using other types of arrays. Since the calculators at GEDmatch are based on the Eurasian ascertained Human Origins dataset to the best of my knowledge, using other arrays ascertained by non W Eurasian samples may lend a new perspective on admixture results.

Also down the road is increasing accuracy in archaic comparisons

Ryukendo
01-16-2016, 02:37 AM
Update

I have not posted anything in the past few days, as I scrapped all my databases and am starting from scratch, with a renewed focus on creating a couple of robust databases, with an emphasis on maximizing markers and overlap with the thousands of samples I have. I am hoping this will increase accuracy, and I can pick up where I left off with Kumtepe.

This is a rather laborious time intensive process, and I anticipate being up and ready in a few days.

In the pipeline is also the idea of branching out of Eurasian ascertained SNP arrays, and doing analysis using other types of arrays. Since the calculators at GEDmatch are based on the Eurasian ascertained Human Origins dataset to the best of my knowledge, using other arrays ascertained by non W Eurasian samples may lend a new perspective on admixture results.

Also down the road is increasing accuracy in archaic comparisons

Kurd, the affymetrix human origins array is not ascertained on eurasians actually. It contains both mbuti, yoruba and san ascertained snps, and all panels are compared against a chimp reference sequence. I do wonder if bias is instead introduced on other arrays with overlap, such as illumina.

Kurd
01-16-2016, 03:02 AM
Kurd, the affymetrix human origins array is not ascertained on eurasians actually. It contains both mbuti, yoruba and san ascertained snps, and all panels are compared against a chimp reference sequence. I do wonder if bias is instead introduced on other arrays with overlap, such as illumina.

I guess I should have clarified. I was referring to the 13 Lazaridis HO panels referenced in the 2012 Patterson-Moorjani paper.

The reason I suspect that Eurasian ascertained panels are more widely used is that Mbutis routinely appear more Neanderthal and Denisovan shifted than Eurasians appear. The panels that show Eurasians more Neanderthal/Denisovan admixed are in fact theSan ascertained panels, but I just don't see much value in a panel that is biased to San derived alleles in a Eurasian to Eurasian study. What do you think?

Ryukendo
01-17-2016, 03:29 PM
Kurd, hmm ok, this might be a basic rookie mistake, but I thought that all 13 panels, incl the African ones, are genotyped on the chip, so ancient and modern genomes when used in this array will capture literally most of the variation that exists across the continents, incl. Africa? Plus one dimension of variation between Humans and Denisova/Neanderthal. Do you mean to say that the ancient genomes we get are genotyped with SNPs on only a few of the panels, and that these are biased towards Eurasians?

Off memory, the original set had something like 800, 000 snps, no? Maybe the reduction in overlap as we end up with ~300 000 snps, as we keep adding genomes to our datasets, eventually cut out all of the non-Eurasian derived snps?

More generally, as we get more stuff out from the Middle East, Africa and S Asia with a lot of 'basal variation', I suspect that getting more African variation might be very useful indeed.

Nick has also replied to an email abt the effect of ascertainment bias on D and F stats, and his confidence on these vs IBS. I have posted his main points in the thread w the Onge, CHG, etc. Just a heads up. I've also invited Shaikorth, Chad and anyone else to help in getting our understanding of these phenomena up to scratch, as we have seen more and more contradictory things recently. We may have to clarify with some scientists abt the things we're seeing, do chip in :)

Padre Organtino
02-08-2016, 06:29 PM
For what to be true, exactly ? For the turnover in the steppe during the Yamnaya -> catacomb -> Srubnaya transition ? No qualifications are required - it happened.

You have Yamnaya rich in CHG and then suddenly you have Andronovo and Sintashta very poor in it but at the same time harboring a lot of Kumtepe. So the continuity between the two groups goes out in the window and Bornze Age Steppe IEs need to derive from a separate mix of EHG and Kumtepe.

IMO, everything speaks for Kumtepe being a phantom of the Admixture like Mozabite or Kalash components.

Gravetto-Danubian
02-08-2016, 08:11 PM
You have Yamnaya rich in CHG and then suddenly you have Andronovo and Sintashta very poor in it but at the same time harboring a lot of Kumtepe. So the continuity between the two groups goes out in the window and Bornze Age Steppe IEs need to derive from a separate mix of EHG and Kumtepe.

IMO, everything speaks for Kumtepe being a phantom of the Admixture like Mozabite or Kalash components.

I was not suggesting Z93 are a wholly different global population. But they did 'replace' earlier Z2103 groups.
Perhaps Z93 came from a more western part of the steppe
So I agree we can't hang too much on this one Kum6 sample- it might mean nothing , and we need more samples