Devils-Gate-Neolithic close modern populations

**Zoro** · 07-02-2020, 04:51 AM

The ancient DNA of two 7700-year-old women from a mountainous cave in far east Russia suggests they were closely related to the people who live in this remote and frigid corner of Asia today. The new discovery also suggests that in this region, farming spread through gradual cultural changes, rather than by an influx of farming people.

Nowadays these genomes are used as proxies for Ancient East Asian ancestry. Out of curiousity I took the highest 2 coverage samples and ran them through Plink's IBS program to see who they are genetically closer to.

I don't have samples from all Siberian populations but out of the samples I have here are the results Devil-Gate sample NEO236:

1- Closest is the ancient Devils-Gate Neo240 sample. From moderns they are closest to modern Japanese followed by other E. Asians.

2- They are furthest from Africans.

3- After E. Asians come northern S. Asians and Central Asian Tajiks.

4- Following these are W. Asians with Kurds and some Caucasians closest and Bedouin, Assyrians and S. Iranians furthest.

5- Following W. Asians come Europeans with Finns on top and Sardinians on bottom.

6- From the highest quality ancient genomes I have Botai is closest and Stuttgart-EEF is furthest.

Over the next few days I'll do similar runs for EEF, WHG, Eastern Yamnaya, Botai and S/SC Asians.

NO	FID1	IID1	FID2	IID2	DISTANCE
1	DevilsCave-N	NEO236	DevilsCave-N	NEO240	0.82125
2	DevilsCave-N	NEO236	Japanese	S-Japanese-1	0.80418
3	DevilsCave-N	NEO236	Japanese	S-Japanese-3	0.80361
4	DevilsCave-N	NEO236	Japanese	S-Japanese-2	0.80333
5	DevilsCave-N	NEO236	Mongola	S-Mongola-2	0.80324
6	DevilsCave-N	NEO236	Han	S-Han-1	0.80306
7	DevilsCave-N	NEO236	Mongola	S-Mongola-1	0.80272
8	DevilsCave-N	NEO236	Yakut	S-Yakut-2	0.80206
9	DevilsCave-N	NEO236	Han	S-Han-2	0.80201
10	DevilsCave-N	NEO236	Buryat	Buryats2	0.80195
11	DevilsCave-N	NEO236	Han	B-Han-3	0.80099
12	DevilsCave-N	NEO236	Yakut	S-Yakut-1	0.80054
13	DevilsCave-N	NEO236	Kalmyk	Kalmyks2	0.80011
14	DevilsCave-N	NEO236	Buryat	Buryats1	0.79995
15	DevilsCave-N	NEO236	Dai	B-Dai-4	0.79974
16	DevilsCave-N	NEO236	Kalmyk	Kalmyks1	0.79930
17	DevilsCave-N	NEO236	Dai	A-Dai-5	0.79923
18	DevilsCave-N	NEO236	Karitiana	BI16	0.79861
19	DevilsCave-N	NEO236	China-Lahu	S-Lahu-2	0.79861
20	DevilsCave-N	NEO236	Dai	S-Dai-1	0.79853
21	DevilsCave-N	NEO236	China-Lahu	S-Lahu-1	0.79827
22	DevilsCave-N	NEO236	Dai	S-Dai-2	0.79790
23	DevilsCave-N	NEO236	Karitiana	S-Karitiana-2	0.79568
24	DevilsCave-N	NEO236	Karitiana	B-Karitiana-3	0.79503
25	DevilsCave-N	NEO236	Karitiana	S-Karitiana-1	0.79496
26	DevilsCave-N	NEO236	Kazakh	Kazakhs2	0.79412
27	DevilsCave-N	NEO236	Kazakh	Kazkahs1	0.79322
28	DevilsCave-N	NEO236	Karakalpak	Karakalpaks2	0.79193
29	DevilsCave-N	NEO236	Kazakh	Kazakhs3	0.79134
30	DevilsCave-N	NEO236	Karakalpak	Karakalpaks3	0.79127
31	DevilsCave-N	NEO236	Karakalpak	Karakalpaks1	0.78937
32	DevilsCave-N	NEO236	Uyghur	Uyghurs2	0.78922
33	DevilsCave-N	NEO236	Turkmen	Turkmens2	0.78824
34	DevilsCave-N	NEO236	Uyghur	Uyghurs1	0.78803
35	DevilsCave-N	NEO236	Uzbek	Uzbeks2	0.78788
36	DevilsCave-N	NEO236	Uyghur	Uyghurs3	0.78778
37	DevilsCave-N	NEO236	Uyghur	S-Uygur-1	0.78744
38	DevilsCave-N	NEO236	Tatar-Tomsk	TomskTatars2	0.78728
39	DevilsCave-N	NEO236	Tatar-Tomsk	TomskTatars1	0.78721
40	DevilsCave-N	NEO236	Uzbek	Uzbeks1	0.78721
41	DevilsCave-N	NEO236	Bashkir	Bashkirs2	0.78658
42	DevilsCave-N	NEO236	Uyghur	S-Uygur-2	0.78591
43	DevilsCave-N	NEO236	Tajik	Tajiks1	0.78323
44	DevilsCave-N	NEO236	Tajik-Tajikistan	Tajik-Tajikistan	0.78311
45	DevilsCave-N	NEO236	Uzbek	Uzbeks3	0.78301
46	DevilsCave-N	NEO236	Turkmen	Turkmens1	0.78266
47	DevilsCave-N	NEO236	Bashkir	Bashkirs1	0.78185
48	DevilsCave-N	NEO236	Bashkir	Bashkirs3	0.78091
49	DevilsCave-N	NEO236	Kapu	S-Kapu-2	0.78073
50	DevilsCave-N	NEO236	Kazakhstan-Botai	BOT15	0.78040
51	DevilsCave-N	NEO236	Kapu	S-Kapu-1	0.77917
52	DevilsCave-N	NEO236	Punjabi	S-Punjabi-2	0.77906
53	DevilsCave-N	NEO236	Punjabi	S-Punjabi-1	0.77904
54	DevilsCave-N	NEO236	Tatar-Volga	VolgaTatars1	0.77875
55	DevilsCave-N	NEO236	Punjabi	S-Punjabi-3	0.77837
56	DevilsCave-N	NEO236	Punjabi-Pk	Punjabi-Pk	0.77820
57	DevilsCave-N	NEO236	Tajik	Tadjik	0.77799
58	DevilsCave-N	NEO236	Tatar-Volga	VolgaTatars2	0.77799
59	DevilsCave-N	NEO236	Tajik	Tajiks3	0.77787
60	DevilsCave-N	NEO236	Punjabi-Pk	Punjabi-Pk	0.77736
61	DevilsCave-N	NEO236	Punjabi-IN	Punjabi-IN	0.77677
62	DevilsCave-N	NEO236	Kalash	S-Kalash-2	0.77647
63	DevilsCave-N	NEO236	Sindhi	S-Sindhi-2	0.77626
64	DevilsCave-N	NEO236	Punjabi-Arain	Punjabi-Arain	0.77601
65	DevilsCave-N	NEO236	Sindhi	S-Sindhi-1	0.77595
66	DevilsCave-N	NEO236	Pashtun-Pak	Pashtun-Pak	0.77580
67	DevilsCave-N	NEO236	Pashtun-Afg	Pashtun-Afg	0.77578
68	DevilsCave-N	NEO236	Punjabi-Pk	Punjabi-Pk	0.77561
69	DevilsCave-N	NEO236	Punjabi-Gujjar	Punjabi-Gujjar	0.77554
70	DevilsCave-N	NEO236	Punjabi	S-Punjabi-4	0.77548
71	DevilsCave-N	NEO236	Pashtun-Afg	Pashtun-Afg	0.77542
72	DevilsCave-N	NEO236	Kurd-Kurmanji-IQ	Kurd-Kurmanji-IQ	0.77533
73	DevilsCave-N	NEO236	Pathan	S-Pathan-1	0.77531
74	DevilsCave-N	NEO236	Pashtun-Pak	Pashtun-Pak	0.77490
75	DevilsCave-N	NEO236	Kalash	S-Kalash-1	0.77472
76	DevilsCave-N	NEO236	Adygei	S-Adygei-1	0.77470
77	DevilsCave-N	NEO236	Pathan	S-Pathan-2	0.77469
78	DevilsCave-N	NEO236	Balochi	S-Balochi-2	0.77434
79	DevilsCave-N	NEO236	Tajik	S-Tajik-1	0.77409
80	DevilsCave-N	NEO236	Brahui	S-Brahui-2	0.77393
81	DevilsCave-N	NEO236	Russia-N-Ossetian	S-North-Ossetian-1	0.77348
82	DevilsCave-N	NEO236	Tajik	S-Tajik-2	0.77344
83	DevilsCave-N	NEO236	Turkish	S-Turkish-2	0.77324
84	DevilsCave-N	NEO236	Kurd-Kurmanji-IQ	Kurd-Kurmanji-IQ	0.77311
85	DevilsCave-N	NEO236	Kurd-Feyli	Kurd-Feyli	0.77300
86	DevilsCave-N	NEO236	Brahui	S-Brahui-1	0.77296
87	DevilsCave-N	NEO236	Kurd-Kurmanji-IQ	Kurd-Kurmanji-IQ	0.77252
88	DevilsCave-N	NEO236	Makrani	S-Makrani-1	0.77233
89	DevilsCave-N	NEO236	Kazakhstan-EBA-Yamnaya	Yamnaya	0.77232
90	DevilsCave-N	NEO236	Iranian	Iranian	0.77232
91	DevilsCave-N	NEO236	Adygei	S-Adygei-2	0.77227
92	DevilsCave-N	NEO236	Turkish	S-Turkish-1	0.77206
93	DevilsCave-N	NEO236	Balochi	S-Balochi-1	0.77204
94	DevilsCave-N	NEO236	Finnish	S-Finnish-1	0.77203
95	DevilsCave-N	NEO236	Finnish	S-Finnish-2	0.77172
96	DevilsCave-N	NEO236	Baloch-IR	Baloch-IR	0.77148
97	DevilsCave-N	NEO236	Turkish	Turkish	0.77147
98	DevilsCave-N	NEO236	Kurd-Feyli	Kurd-Feyli	0.77131
99	DevilsCave-N	NEO236	Iranian	S-Iranian-2	0.77126
100	DevilsCave-N	NEO236	Iranian	S-Iranian-1	0.77125
101	DevilsCave-N	NEO236	Estonian	S-Estonian-2	0.77121
102	DevilsCave-N	NEO236	Kurd-Turkey	Kurd-Turkey	0.77094
103	DevilsCave-N	NEO236	Kurd-Feyli	Kurd-Feyli	0.77074
104	DevilsCave-N	NEO236	Makrani	S-Makrani-2	0.77070
105	DevilsCave-N	NEO236	Lezgin	S-Lezgin-1	0.77066
106	DevilsCave-N	NEO236	Lezgin	S-Lezgin-2	0.77062
107	DevilsCave-N	NEO236	Armenian	S-Armenian-1	0.77053
108	DevilsCave-N	NEO236	Georgian	S-Georgian-2	0.77008
109	DevilsCave-N	NEO236	Baloch-IR	Baloch-IR	0.76992
110	DevilsCave-N	NEO236	Kurd-Feyli	Kurd-Feyli	0.76988
111	DevilsCave-N	NEO236	Finnish	S-Finnish-3	0.76986
112	DevilsCave-N	NEO236	Georgian	S-Georgian-1	0.76955
113	DevilsCave-N	NEO236	English	S-English-1	0.76944
114	DevilsCave-N	NEO236	Estonian	S-Estonian-1	0.76941
115	DevilsCave-N	NEO236	Armenian	S-Armenian-2	0.76938
116	DevilsCave-N	NEO236	Hungarian	S-Hungarian-1	0.76907
117	DevilsCave-N	NEO236	Iranian	Iranian	0.76906
118	DevilsCave-N	NEO236	Basque	S-Basque-1	0.76888
119	DevilsCave-N	NEO236	Assyrian	Assyrian	0.76876
120	DevilsCave-N	NEO236	Assyrian	Assyrian	0.76875
121	DevilsCave-N	NEO236	Hungarian	S-Hungarian-2	0.76871
122	DevilsCave-N	NEO236	Spanish	S-Spanish-1	0.76857
123	DevilsCave-N	NEO236	Spanish	S-Spanish-2	0.76829
124	DevilsCave-N	NEO236	Basque	S-Basque-2	0.76828
125	DevilsCave-N	NEO236	English	S-English-2	0.76800
126	DevilsCave-N	NEO236	BedouinB	S-BedouinB-2	0.76752
127	DevilsCave-N	NEO236	Sardinian	S-Sardinian-2	0.76716
128	DevilsCave-N	NEO236	Sardinian	S-Sardinian-1	0.76702
129	DevilsCave-N	NEO236	Loschbour-WHG	Loschbour-snpAD	0.76663
130	DevilsCave-N	NEO236	Sardinian	B-Sardinian-3	0.76662
131	DevilsCave-N	NEO236	Stuttgart-EEF	Stuttgart-published	0.76585
132	DevilsCave-N	NEO236	BedouinB	S-BedouinB-1	0.76549
133	DevilsCave-N	NEO236	Somali	S-Somali-1	0.74463
134	DevilsCave-N	NEO236	Masai	S-Masai-1	0.73893
135	DevilsCave-N	NEO236	Masai	S-Masai-2	0.73745
136	DevilsCave-N	NEO236	Luhya	S-Luhya-2	0.72884
137	DevilsCave-N	NEO236	Luhya	S-Luhya-1	0.72868
138	DevilsCave-N	NEO236	Mbuti	S-Mbuti-3	0.72837
139	DevilsCave-N	NEO236	Yoruba	S-Yoruba-1	0.72825
140	DevilsCave-N	NEO236	Mbuti	S-Mbuti-2	0.72789
141	DevilsCave-N	NEO236	Yoruba	S-Yoruba-2	0.72671
142	DevilsCave-N	NEO236	Yoruba	B-Yoruba-3	0.72566

**Zoro** · 07-02-2020, 03:15 PM

Out of curiosity I went to Vahuado site to see if the G25 results also make sense for Devils-Gate-N. Unfortunately the G25 results were NONSENSICAL as you can see below.

You don't have to take my word for it. Just go to Vahuado and copy Devils-Gate from G25 SCALED AVG Ancient spreadsheet and paste it in TARGET in G25 SCALED AVG Moderns calculator.

Hopefully this will convince those who are still on the fence to not take the G25 outputs seriously.

Here are some G25 nonsensical results compared to my PLINK IBS results above. You maybe able to spot even more illogical results.

LIST OF G25 NONSENSICAL RESULTS
NO	NONSENSICAL RESULT	IBS- PLINK	DAVIDSKI – G25
1	Devils-Gate closer to many S.& W Asians & Europeans than to E. Eurasians	NO	YES
2	Devils-Gate closer to Libyans, Algerians, Egyptians, Berber than to Lithuanians & Armenians	NO	YES
3	Devils-Gate closer to S. Iranians than to Kurds	NO	YES

G25 Modern Averages scaled, official datasheet
POPULATION	Distance to RUS_Devils_Gate_Cave_N
Mongola	0.09648
Korean	0.12362
Japanese	0.12374
Tibetan_Gangcha	0.12467
Kalmyk	0.13047
Mongolian	0.13589
Buryat	0.13749
Han_Jiangsu	0.15949
Han_Shanghai	0.16224
Han_Zhejiang	0.16848
Han_Hubei	0.17012
Han_Sichuan	0.17243
Koryak	0.18688
Yakut	0.19216
Dai	0.24433
Kazakh	0.24489
Brahmin_Manipuri	0.28047
Karakalpak	0.28056
Uygur	0.31489
Uzbek	0.36435
Bashkir	0.37197
Turkmen	0.42475
Gond	0.42698
Paniya	0.45062
Irula	0.45444
Pulliyar	0.45849
Burusho	0.46823
North_Kannadi	0.46981
Hakkipikki	0.47077
Uttar_Pradesh	0.47749
Mala	0.47761
Yadava	0.47845
Maratha	0.47910
Dusadh	0.47939
Kol	0.48416
Tatar_Kazan	0.48508
Piramalai	0.48621
Kanjar	0.48656
Punjabi_Lahore	0.48694
Mixtec	0.48810
Zapotec	0.48925
Pima	0.48936
Velamas	0.49030
Brahmin_Tamil_Nadu	0.49132
Tajik	0.49203
Gujarati	0.49562
Kashmiri_Pandit	0.49853
Mayan	0.49890
Mixe	0.49892
Quechua	0.50299
Brahmin_Gujarat	0.50491
Brahmin_Uttar_Pradesh	0.50595
Tatar_Mishar	0.51283
Pashtun	0.51377
Tajik_Ishkashim	0.51603
Gujar_Pakistan	0.51798
Sindhi	0.51821
Punjabi_Jatt	0.51933
Pashtun-Yusufzai	0.52017
Pashtun-Uthmankhel	0.52148
Tajik_Shugnan	0.52207
Bolivian_LaPaz	0.52214
Pashtun-Tarkalani	0.52294
Tajik_Rushan	0.52785
Kalash	0.52873
Brahui	0.54727
Balochi	0.54963
Parsi_India	0.55047
Parsi_Pakistan	0.55150
Makrani	0.55650
Circassian	0.55678
Iranian_Bandari	0.55968
North_Ossetian	0.56144
Finnish_East	0.56274
Karitiana	0.56404
Iranian_Fars	0.57057
Surui	0.57266
Finnish	0.57687
Adygei	0.57706
Australian	0.57981
Iranian_Lor	0.58065
Tabasaran	0.58172
Iranian_Zoroastrian	0.58288
Kurdish	0.58813
Syrian	0.59025
Abkhasian	0.59109
Jordanian	0.59519
Russian_Kursk	0.59637
Moldovan	0.59668
Libyan	0.59770
BedouinA	0.59798
Russian_Orel	0.59873
Egyptian	0.59881
Algerian	0.60187
Estonian	0.60205
Berber_Tunisia_Sen	0.60223
Hungarian	0.60227
Belarusian	0.60644
Albanian	0.60646
Armenian	0.60886
Spanish_Asturias	0.61105
Spanish_Andalucia	0.61165
Lithuanian_RA	0.61167
Spanish_Galicia	0.61185
Spanish_Soria	0.61193
English	0.61247
Lithuanian_SZ	0.61545
Sardinian	0.63860
Somali	0.65247
Masai	0.71578
Papuan	0.75843
Ethiopian_Gumuz	0.75873
Ethiopian_Mursi	0.76359
Ethiopian_Anuak	0.79354
Luhya_Kenya	0.80931
Yoruba	0.83579
Mbuti	1.00219

**vbnetkhio** · 07-02-2020, 03:22 PM

Originally Posted by Zoro

The ancient DNA of two 7700-year-old women from a mountainous cave in far east Russia suggests they were closely related to the people who live in this remote and frigid corner of Asia today. The new discovery also suggests that in this region, farming spread through gradual cultural changes, rather than by an influx of farming people.

Nowadays these genomes are used as proxies for Ancient East Asian ancestry. Out of curiousity I took the highest 2 coverage samples and ran them through Plink's IBS program to see who they are genetically closer to.

I don't have samples from all Siberian populations but out of the samples I have here are the results Devil-Gate sample NEO236:

1- Closest is the ancient Devils-Gate Neo240 sample. From moderns they are closest to modern Japanese followed by other E. Asians.

2- They are furthest from Africans.

3- After E. Asians come northern S. Asians and Central Asian Tajiks.

4- Following these are W. Asians with Kurds and some Caucasians closest and Bedouin, Assyrians and S. Iranians furthest.

5- Following W. Asians come Europeans with Finns on top and Sardinians on bottom.

6- From the highest quality ancient genomes I have Botai is closest and Stuttgart-EEF is furthest.

Over the next few days I'll do similar runs for EEF, WHG, Eastern Yamnaya, Botai and S/SC Asians.

is IBS similar to FST?

**Zoro** · 07-02-2020, 03:35 PM

Originally Posted by vbnetkhio

is IBS similar to FST?

yes it's similar but I think that the plink --genome flag is more useful because it also outputs IBD. Pi-Hat, Z2, Z1, Z0 and PPC are really useful if you're trying to figure out more recent admixture

It's pretty simple you can use:

/plink --bfile test --genome

**Zoro** · 07-02-2020, 03:48 PM

Based on the above G25 illogical results for Devils-Gate-N how can anyone trust the G25 ancient models or distances

**Kyp** · 07-02-2020, 03:55 PM

Target: FatherKyp
Distance: 1.5896% / 0.01589648
31.6 Anatolia_Barcin_N
23.8 IRN_Ganj_Dareh_N
18.8 Yamnaya_RUS_Samara
10.4 RUS_Devils_Gate_Cave_N
8.6 GEO_CHG
6.8 Levant_Natufian

**Zoro** · 07-02-2020, 04:05 PM

Originally Posted by Kyp Snow

Target: FatherKyp
Distance: 1.5896% / 0.01589648
31.6 Anatolia_Barcin_N
23.8 IRN_Ganj_Dareh_N
18.8 Yamnaya_RUS_Samara
10.4 RUS_Devils_Gate_Cave_N
8.6 GEO_CHG
6.8 Levant_Natufian

I expect your dad to score Devils-Gate-N because that't the ancient proxy but what I'm saying is that if the G25 distances are screwed up for Devils-Gate (above table) then it's likely the distances for the other ancients are screwed up too in the G25 which means the percentages in your dad's model are screwed up.

I'm planning to do a comparison of G25 against Plink IBS for some of the other high quality ancients such as Stuttgart, Loschbour, Botai, Yamnaya so we can have a better idea how screwed up the G25 is. I remember the nMonte designer Hudjbrets wrote in the readme file Garbage in --> Garbage out lol.

**vbnetkhio** · 07-02-2020, 04:11 PM

Originally Posted by Zoro

Based on the above G25 illogical results for Devils-Gate-N how can anyone trust the G25 ancient models or distances

the proper tools are too complicated for the average user and require a lot of ram and processing power

**Zoro** · 07-02-2020, 04:24 PM

Originally Posted by vbnetkhio

the proper tools are too complicated for the average user and require a lot of ram and processing power

I agree but it’s annoying to see how users who don’t know better blindly put faith in Davidski G25 as if it’s the gospel when we can see it’s quite off.

Actually plink by itself without using Admixture doesn’t take much time to process and a good dataset is only about 1 gig

Edit: It would get complicated as you say if they have to use Admixtools to convert dataset to plink but if they can find a dataset in bed bim fam format and only use plink then it’s not too bad but they would have to know how to do QC on dataset and work with Linux

**vbnetkhio** · 07-02-2020, 04:42 PM

Originally Posted by Zoro

I agree but it’s annoying to see how users who don’t know better blindly put faith in Davidski G25 as if it’s the gospel when we can see it’s quite off.

Actually plink by itself without using Admixture doesn’t take much time to process and a good dataset is only about 1 gig

but in plink you can't use outgroups, right?
i'm interested in comparing BA and IA samples to moderns, and qpAdm is perfect for this.
because i can include that ancient stuff like WHG, EEF etc. as outgroups, and then the algorithm actually comapres the BA onwards drift