PDA

View Full Version : New High-Coverage Northeast African genomes



Awale
08-25-2017, 08:24 PM
Northeast African genomic variation shaped by the continuity of indigenous groups and Eurasian migrations (http://journals.plos.org/plosgenetics/article?id=10.1371%2Fjournal.pgen.1006976)

Abstract:


Northeast Africa has a long history of human habitation, with fossil-finds from the earliest anatomically modern humans, and housing ancient civilizations. The region is also the gate-way out of Africa, as well as a portal for migration into Africa from Eurasia via the Middle East and the Arabian Peninsula. We investigate the population history of northeast Africa by genotyping ~3.9 million SNPs in 221 individuals from 18 populations sampled in Sudan and South Sudan and combine this data with published genome-wide data from surrounding areas. We find a strong genetic divide between the populations from the northeastern parts of the region (Nubians, central Arab populations, and the Beja) and populations towards the west and south (Nilotes, Darfur and Kordofan populations). This differentiation is mainly caused by a large Eurasian ancestry component of the northeast populations likely driven by migration of Middle Eastern groups followed by admixture that affected the local populations in a north-to-south succession of events. Genetic evidence points to an early admixture event in the Nubians, concurrent with historical contact between North Sudanese and Arab groups. We estimate the admixture in current-day Sudanese Arab populations to about 700 years ago, coinciding with the fall of Dongola in 1315/1316 AD, a wave of admixture that reached the Darfurian/Kordofanian populations some 400–200 years ago. In contrast to the northeastern populations, the current-day Nilotic populations from the south of the region display little or no admixture from Eurasian groups indicating long-term isolation and population continuity in these areas of northeast Africa.

So the headcount is:

44 Sudanese Arabs, 27 Beja, 41 Nubians, 52 Eastern Sudanic speakers (xNubians), 22 Saharan speakers, 14 Copts, 16 Nubas and 5 Hausas.

Some figures:

http://i.imgur.com/Xfd4Ayb.png

http://i.imgur.com/FI7VOrg.png

And the samples are available here (http://jakobssonlab.iob.uu.se/data/).

blackflash16
08-26-2017, 03:15 AM
However, these old admixture events into the Beni Amer could be driven by admixture from the Cushitic-speaking populations of the Horn of Africa, which themselves have received 30–50% non-African ancestry about 100 generations ago, or 3kya.


I wonder if this 3kya nonsense will stop when the Skoglund paper with that ~3000 year old Tanzanian pastoralist is released.

Awale
08-26-2017, 11:25 AM
I wonder if this 3kya nonsense will stop when the Skoglund paper with that ~3000 year old Tanzanian pastoralist is released.

Yes, I think it'll dawn on them when they see a 3kya old ancient genome south of the Horn with probably the same old West-Eurasian admixture as Somalis. Maybe then they'll realize that their ~3kya signal is probably just them picking up on admixture from the Proto-Ethiosemitic speaking community.

nee4speed111
08-27-2017, 01:02 AM
Good study, I'll give it a more thorough read when I get the time, but I'll try to address the main points in it for now. The Figure 2 graph is particularly interesting, it seems to imply that Nilotes have remained relatively unmixed with Eurasian populations, indicating strong genetic isolation of Nilotes.

Also something that I found that was very interesting, was the map that showed the amount of Nilotic and Eurasian admixture and admixture dates in investigated populations. The map doesn't make it very clear, but the S7 table has a breakdown of the relevant dates of admixture for each population sampled http://journals.plos.org/plosgenetic...n.1006976.s035

Most interesting thing to me, is that all the populations sampled seem to have at least one admixture event, sometimes 2. The 700 year figure for Nubians seems to be an average based on the admixture dates of the Mahas, Danagla, Halfawieen, Gaalien & other Nubian tribes. Some of them like the Gaalien have quite recent admixture dates, at approx 17 generations ago or 425 years ago. Others like the Mahas have admixture dates of 58 generations ago, which is approximately 1450 years ago I believe.

The Beja groups are similarly interesting, the Beni Amer have 2 admixture events, one that occurs at 107 generations ago or 2675 years ago, long before any arab or muslim migration, & another admix event that occurs 34 generations ago, or 850 years ago which is probably in relation to arab/muslim migrations

This could be due to admixture from Cushitic groups residing in the Horn at the time, as the date of admixture is very similar to Somali dates of Eurasian admix.

Another interesting piece of data in regards to the admixture dates, is that the only 2 groups who the software doesn't pick up an admix date between the Nilotic/East African & Eurasian admix is Dinka & Copts. I'm still trying to puzzle out the full implications of this particular piece of data, but I've wanted someone to test admix dates in Copts for a while & the fact that the software couldn't pick up a date for either Dinka or Copts have interesting implications, about the genetic history of either group.

Shaikorth
08-27-2017, 11:59 AM
Good study, I'll give it a more thorough read when I get the time, but I'll try to address the main points in it for now. The Figure 2 graph is particularly interesting, it seems to imply that Nilotes have remained relatively unmixed with Eurasian populations, indicating strong genetic isolation of Nilotes.

Also something that I found that was very interesting, was the map that showed the amount of Nilotic and Eurasian admixture and admixture dates in investigated populations. The map doesn't make it very clear, but the S7 table has a breakdown of the relevant dates of admixture for each population sampled http://journals.plos.org/plosgenetic...n.1006976.s035

Most interesting thing to me, is that all the populations sampled seem to have at least one admixture event, sometimes 2. The 700 year figure for Nubians seems to be an average based on the admixture dates of the Mahas, Danagla, Halfawieen, Gaalien & other Nubian tribes. Some of them like the Gaalien have quite recent admixture dates, at approx 17 generations ago or 425 years ago. Others like the Mahas have admixture dates of 58 generations ago, which is approximately 1450 years ago I believe.

The Beja groups are similarly interesting, the Beni Amer have 2 admixture events, one that occurs at 107 generations ago or 2675 years ago, long before any arab or muslim migration, & another admix event that occurs 34 generations ago, or 850 years ago which is probably in relation to arab/muslim migrations

This could be due to admixture from Cushitic groups residing in the Horn at the time, as the date of admixture is very similar to Somali dates of Eurasian admix.

Another interesting piece of data in regards to the admixture dates, is that the only 2 groups who the software doesn't pick up an admix date between the Nilotic/East African & Eurasian admix is Dinka & Copts. I'm still trying to puzzle out the full implications of this particular piece of data, but I've wanted someone to test admix dates in Copts for a while & the fact that the software couldn't pick up a date for either Dinka or Copts have interesting implications, about the genetic history of either group.

Note that they tested admix dates with LD-based methods which are easily confused by drift of endogamous groups. ALDER fails to find admixture dates for Kalash (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4570283/). That said, Dinka do not look very drifted and likely just don't have European or Near Eastern ancestry.

TuaMan
08-27-2017, 01:58 PM
At K=4 it looks like Nuer, Dinka, and Anuak are the three Nilotic populations completely lacking in the CEU/European component (Shilluk looks like it just barely has some trace amounts). They all have a solid chunk of the orange Yoruba component though. Has West African ancestry been corroborated in any of these groups before? And Yoruba has an even bigger chunk of the dark blue Nilotic-esque component, which is pretty interesting.

Awale
08-27-2017, 05:27 PM
The Beja groups are similarly interesting, the Beni Amer have 2 admixture events, one that occurs at 107 generations ago or 2675 years ago, long before any arab or muslim migration ... This could be due to admixture from Cushitic groups residing in the Horn at the time, as the date of admixture is very similar to Somali dates of Eurasian admix.

I half wonder if it's actually tied to the Proto-Ethiopiansemitic speaking community as well given the timing. It's not impossible for them to have been influenced by it as well. Lately I've been wondering if all of the later West-Eurasian admixture in the Horn (http://anthromadness.blogspot.ae/2016/01/the-later-west-eurasian-element-in-agaws.html) is actually tied to that and that there isn't one specific to Agaws and the like who might just have less of this admixture because they didn't fully language-shift and mix that much with fully Ethiosemitized Horn-Africans. Hopefully, nMonte might help get to the bottom of this if and when we can get all these samples in the global 10 PCA and perhaps also a more regional one like this one (https://drive.google.com/file/d/0B2ARnUeK-Y8WS2JfNGYzelpZODQ/view). Then we might be able see if there's a pulse of something more Copt-like before the Yemenite Jew-like stuff.


Has West African ancestry been corroborated in any of these groups before?

Yes, that's been known for a long time now. The usual Dinka samples out there before these having, on average, a bit less of such admixture than the usual Ethiopian-Anuaks and Sudanese samples is probably why they work better as a proxy for Somalis' African ancestry, in fact:

[1] "distance%=0.6546 / distance=0.006546"

Somali

Dinka 54.6
Natufian 42.6
MA1 2.3
Mota 0.4

--

[1] "distance%=0.95 / distance=0.0095"

Somali

Ethiopian_Anuak 53.05
Natufian 45.30
Ari-Blacksmith 0.70
MA1 0.60
Mota 0.35

--

[1] "distance%=1.0303 / distance=0.010303"

Somali

Sudanese 53.5
Natufian 45.8
MA1 0.8

TuaMan
08-27-2017, 08:15 PM
Yes, that's been known for a long time now. The usual Dinka samples out there before these having, on average, a bit less of such admixture than the usual Ethiopian-Anuaks and Sudanese samples is probably why they work better as a proxy for Somalis' African ancestry, in fact:

Interesting. Uniparentally, I know Nilo-Saharan types are dominated by y-DNA A and B and I believe maternally are predominantly L3. So do they get their West-African ancestry from their maternal side (I believe they also have some L2 mtDNA, correct me if I'm wrong).

pgbk87
08-27-2017, 09:19 PM
Some of my observations:


Based on ADMIXTURE, the Nuba and the Nuer look to be the "purest" Nilotes. I wonder how they would behave as a proxy for Somali's African ancestry
The Great Lakes Bantu groups seem to have significant Somali-like and Anuak-like admixture
The Beja (Beni Amer) have a small Omotic-like signal at K=10, similar to Somalis, but it disappears at later Ks
Most of the Sudanese Arabs have less of a West African-like admixture signal than Horn Africans(?), minus a few outliers
At K=18 the Yoruba have a small Kordofan/Darfurian component that makes sense linguistically
At K=16 and 18 there is a Sudanese centered component that simply does not exist in Nilo-Saharans proper, and maximizes in the Danagla
If you look closely, there is tiny tan-colored component that I'd guess is something hunter-gather-esque in the Ari Cultivators, Anuak, Great Lakes and South Sudanese groups

Angoliga
08-27-2017, 11:17 PM
I'd love to see puntDNAL work with this data for a new admix calc :). Hopefully there'll be something of value for him with these new samples and components.


Relating to Nubians, I found the dating within these confidence intervals relevant to the decline of the Kingdom of MeroŽ (1st-4th century AD):

*If the LD-decay estimates have any merit, could these Eurasian introgression dates for the Danagla and Mahas reflect the turning point of the Nobatae, Blemmyes (https://en.wikipedia.org/wiki/Blemmyes), and later Aksumites (https://en.wikipedia.org/wiki/Kingdom_of_Aksum)conquering Nubia?

The roughly 56 generation estimate (56x30=~1680ybp) happens to fall within this time period:


"Assuming that the Nubian population is a mixture of an incoming Eurasian (TSI is used as a proxy) group and a resident group that is genetically similar to the current day Nilotes (Nuer is used as a proxy), first contact is dated using patterns of LD-decay [34] to roughly 56 generations ago for the Danagla (54.45 +/- 10.34, Z = 5.26437) and the Mahas (58.35 +/- 12.2, Z = 4.78402); the Halfawieen have received Eurasian admixture later, around 19 generations ago (19.31 +/- 3.81, Z = 5.05949, S7 Table, Fig 3C). Assuming a generation time of 30 years, the admixture dates for Danagla and Mahas predate the Arab expansion in the 7th century, and may suggest that the migrations and admixture predate Islamic conquest. However, the confidence intervals overlap with the 7th century, and these admixture estimates largely coincide with the Arab expansion into the northeast of Sudan. It is known from historic sources that Arabic groups encountered the Nubians first in the 7th century, and were held back from advancing further into the Sahel until the fall of Dongola in 1315/1316AD [36] and the collapse of the Kingdom of Makuria. This is consistent with the later date for the admixture into Halfawieen and the Arabic populations of Sudan. Previous studies [37, 38] have found a similar pattern for populations of Maghreb, where admixture times coincide with the time of the historically documented Arab conquest."

This would agree with the Sudanes YDNA from ancient remains:


"Accordingly, through limited on number of aDNA samples, there is enough data to suggest and to tally with the historical evidence of the dominance by Nilotic elements during the early state formation in the Nile Valley, and as the states thrived there was a dominance by other elements particularly Nuba/Nubians. In Y-chromosome terms this mean in simplest terms introgression of the YAP insertion (haplogroups E and D), and Eurasian Haplogroups which are defined by F-M89 against a background of haplogroup A-M13. The data analysis of the extant Y-chromosomes suggests that the bulk of genetic diversity appears to be a consequence of recent migrations and demographic events" (Hisham, 2009 (http://khartoumspace.uofk.edu/handle/123456789/6376)).


Brief summary (https://en.wikipedia.org/wiki/Roman%E2%80%93Nubian_relations#History_of_Rome_.26 _Nubia)of the decline of MeroŽ (1st-4th century):


"..the Kingdom of Meroe began to fade as a power by the 1st or 2nd century AD, sapped by the war with Roman Egypt, the decline of its traditional industries and the increasingly destructive attacks from the Blemmyes.

Indeed, at the end of the 2nd century AD, the Nobatae occupied the Nile's west bank in northern Kush. They are believed to have been one of several well-armed bands of horse- and camel-borne warriors who sold their
vagility to the Meroitic Population for protection; eventually they intermarried and established themselves among the Meroitic people as a military aristocracy.

...the old Meroitic kingdom contracted because of the expansion of the powerful Ethiopic Kingdom of Aksum to the east. By AD 350, King Ezana of Axum
had captured and destroyed Meroe city, ending the kingdom's independent existence, and conquering its territory into modern-day northern Sudan."

Why no Central Sudanic (https://en.wikipedia.org/wiki/Central_Sudanic_languages)s?


http://i.imgur.com/2UhIpPL.png
Central Sudanic Linguistic Map

This paper had me super elated and depressed at the same time. I was hoping to finally see a true reflection of my own admixture through samples from Central-Sudanic (https://en.wikipedia.org/wiki/Central_Sudanic_languages) speakers but it looks like we've been scapegoated yet again. IMO at least the inclusion of the CS Lugbara (https://en.wikipedia.org/wiki/Lugbara_language)speakers with a population size (~1.7 million) comparable to Nubians, and exceeding many on the list, should have been on the sample-collection-radar. So far, the Chadian Bulala (https://en.wikipedia.org/wiki/Bilala_people)are the only CS samples I've ever come across. Despite our geographical distance (<2,000km - Central Chad to NW Uganda), and our supposed period of migration towards the end of the Holocene Wet Phase (~3500-3000), the Bulala still appear as my closest oracle population. As the paper suggests, many of these NE nilotic populations must have stayed "strongly endogamous" - perhaps for several millennia.

Based on Bulala samples and my own observation with immediate family members, if CS populations (minus pygmy speakers) were included in the paper, I'd assume ~40% of their admixture would be the orange West-African component. The rest would be the blue NE component along with a minor Eurasian affinity (~ <10%) at lower K levels. The closest comparable sample with these fractions are found in one of the Zagawa (https://en.wikipedia.org/wiki/Zaghawa_people) (perhaps an anomaly from neighbouring CS speakers in Chad?).

I'm looking forward to learn more from this thread, I'll definitely chime in on more observations of interest -- these were just the first to come to mind.


*This caught me a little off guard, I think the naming for the nilotic population "Baria" might have been incorrect. This is probably referring to the "Bari (https://en.wikipedia.org/wiki/Bari_people)" of equatorial Southern-Sudan (~1,000,000 pop.).

blackflash16
08-27-2017, 11:44 PM
I merged this dataset with the one used for Lazaridis et al. 2016. I want to make sure I did it properly. Do these d-stats look okay? I bolded the samples from this study.



result: Levant_N Yoruba Nuba Chimp 0.0513 12.115 9610 8672 131084
result: Levant_N Yoruba Dinka Chimp 0.0318 7.637 9308 8734 131072
result: Levant_N Yoruba Somali Chimp 0.1651 40.055 11424 8186 131153
result: Levant_N Mota Somali Chimp 0.1589 29.671 11180 8114 131130
result: Levant_N Mota Nuba Chimp 0.0633 11.258 9532 8398 131062


qpAdm Somali = Nuba + Levant_N:
left pops:
Somali
Nuba
Levant_N

right pops:
Chukchi
Han
Karitiana
Kostenki14
MA1
Mbuti
Papuan
Ust_Ishim

0 Somali 13
1 Nuba 16
2 Levant_N 13
3 Chukchi 23
4 Han 43
5 Karitiana 12
6 Kostenki14 1
7 MA1 1
8 Mbuti 10
9 Papuan 14
10 Ust_Ishim 1
jackknife block size: 0.050
snps: 177419 indivs: 147
number of blocks for block jackknife: 705
## ncols: 177419
coverage: Somali 177419
coverage: Nuba 177320
coverage: Levant_N 134293
coverage: Chukchi 177419
coverage: Han 177419
coverage: Karitiana 177419
coverage: Kostenki14 163879
coverage: MA1 126588
coverage: Mbuti 177419
coverage: Papuan 177419
coverage: Ust_Ishim 177094
dof (jackknife): 606.064
numsnps used: 93145
codimension 1
f4info:
f4rank: 1 dof: 6 chisq: 5.170 tail: 0.522145558 dofdiff: 8 chisqdiff: -5.170 taildiff: 1
B:
scale 1.000
Han 0.096
Karitiana -0.085
Kostenki14 -0.310
MA1 -0.286
Mbuti 2.597
Papuan 0.215
Ust_Ishim 0.126
A:
scale 153.534
Nuba 0.726
Levant_N -1.214


full rank
f4info:
f4rank: 2 dof: 0 chisq: 0.000 tail: 1 dofdiff: 6 chisqdiff: 5.170 taildiff: 0.522145558
B:
scale 1.000 1.000
Han 0.095 0.069
Karitiana -0.086 -1.314
Kostenki14 -0.315 1.606
MA1 -0.285 0.400
Mbuti 2.597 1.194
Papuan 0.213 -0.789
Ust_Ishim 0.123 -0.692
A:
scale 152.302 4496.400
Nuba 0.712 0.474
Levant_N -1.222 1.332


best coefficients: 0.626 0.374
Jackknife mean: 0.625710382 0.374289618
std. errors: 0.006 0.006

error covariance (* 1000000)
42 -42
-42 42


summ: Somali 2 0.522146 0.626 0.374 42 -42 42

fixed pat wt dof chisq tail prob
00 0 6 5.170 0.522146 0.626 0.374
01 1 7 2749.849 0 1.000 -0.000
10 1 7 2163.902 0 0.000 1.000
best pat: 00 0.522146 - -
best pat: 01 0 chi(nested): 2744.679 p-value for nested model: 0

coeffs: 0.626 0.374

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Nuba Han 0.000448 3.750642
details: Levant_N Han -0.000732 -2.892074
dscore: Han f4: 0.000006 Z: 0.054343

details: Nuba Karitiana -0.000526 -2.898164
details: Levant_N Karitiana 0.000385 0.975369
dscore: Karitiana f4: -0.000185 Z: -1.014942

details: Nuba Kostenki14 -0.001234 -4.235744
details: Levant_N Kostenki14 0.003257 4.942878
dscore: Kostenki14 f4: 0.000447 Z: 1.529497

details: Nuba MA1 -0.001243 -4.014649
details: Levant_N MA1 0.002483 3.935602
dscore: MA1 f4: 0.000151 Z: 0.522370

details: Nuba Mbuti 0.012281 46.658528
details: Levant_N Mbuti -0.020438 -40.352990
dscore: Mbuti f4: 0.000034 Z: 0.144178

details: Nuba Papuan 0.000959 4.262129
details: Levant_N Papuan -0.001837 -4.173768
dscore: Papuan f4: -0.000088 Z: -0.432495

details: Nuba Ust_Ishim 0.000502 1.738664
details: Levant_N Ust_Ishim -0.001163 -2.007938
dscore: Ust_Ishim f4: -0.000121 Z: -0.469780

gendstat: Chukchi Han 0.054
gendstat: Chukchi Karitiana -1.015
gendstat: Chukchi Kostenki14 1.529
gendstat: Chukchi MA1 0.522
gendstat: Chukchi Mbuti 0.144
gendstat: Chukchi Papuan -0.432
gendstat: Chukchi Ust_Ishim -0.470
gendstat: Han Karitiana -1.009
gendstat: Han Kostenki14 1.515
gendstat: Han MA1 0.488
gendstat: Han Mbuti 0.122
gendstat: Han Papuan -0.516
gendstat: Han Ust_Ishim -0.509
gendstat: Karitiana Kostenki14 1.969
gendstat: Karitiana MA1 1.051
gendstat: Karitiana Mbuti 0.791
gendstat: Karitiana Papuan 0.391
gendstat: Karitiana Ust_Ishim 0.224
gendstat: Kostenki14 MA1 -0.785
gendstat: Kostenki14 Mbuti -1.223
gendstat: Kostenki14 Papuan -1.769
gendstat: Kostenki14 Ust_Ishim -1.608
gendstat: MA1 Mbuti -0.339
gendstat: MA1 Papuan -0.733
gendstat: MA1 Ust_Ishim -0.758
gendstat: Mbuti Papuan -0.484
gendstat: Mbuti Ust_Ishim -0.531
gendstat: Papuan Ust_Ishim -0.121

## end of run

qpAdm Somali = Nuer + Levant_N:

left pops:
Somali
Nuer
Levant_N

right pops:
Chukchi
Han
Karitiana
Kostenki14
MA1
Mbuti
Papuan
Ust_Ishim

0 Somali 13
1 Nuer 16
2 Levant_N 13
3 Chukchi 23
4 Han 43
5 Karitiana 12
6 Kostenki14 1
7 MA1 1
8 Mbuti 10
9 Papuan 14
10 Ust_Ishim 1
jackknife block size: 0.050
snps: 177419 indivs: 147
number of blocks for block jackknife: 705
## ncols: 177419
coverage: Somali 177419
coverage: Nuer 177323
coverage: Levant_N 134293
coverage: Chukchi 177419
coverage: Han 177419
coverage: Karitiana 177419
coverage: Kostenki14 163879
coverage: MA1 126588
coverage: Mbuti 177419
coverage: Papuan 177419
coverage: Ust_Ishim 177094
dof (jackknife): 606.130
numsnps used: 93114
codimension 1
f4info:
f4rank: 1 dof: 6 chisq: 7.356 tail: 0.289199006 dofdiff: 8 chisqdiff: -7.356 taildiff: 1
B:
scale 1.000
Han 0.097
Karitiana -0.075
Kostenki14 -0.269
MA1 -0.276
Mbuti 2.603
Papuan 0.211
Ust_Ishim 0.122
A:
scale 143.899
Nuer 0.838
Levant_N -1.139


full rank
f4info:
f4rank: 2 dof: 0 chisq: 0.000 tail: 1 dofdiff: 6 chisqdiff: 7.356 taildiff: 0.289199006
B:
scale 1.000 1.000
Han 0.097 0.061
Karitiana -0.075 -0.922
Kostenki14 -0.275 1.997
MA1 -0.275 0.476
Mbuti 2.603 0.923
Papuan 0.210 -0.821
Ust_Ishim 0.121 -0.638
A:
scale 143.053 3538.667
Nuer 0.827 0.395
Levant_N -1.147 1.358


best coefficients: 0.576 0.424
Jackknife mean: 0.576287430 0.423712570
std. errors: 0.007 0.007

error covariance (* 1000000)
42 -42
-42 42


summ: Somali 2 0.289199 0.576 0.424 42 -42 42

fixed pat wt dof chisq tail prob
00 0 6 7.356 0.289199 0.576 0.424
01 1 7 3520.580 0 1.000 -0.000
10 1 7 2157.555 0 0.000 1.000
best pat: 00 0.289199 - -
best pat: 01 0 chi(nested): 3513.225 p-value for nested model: 0

coeffs: 0.576 0.424

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Nuer Han 0.000567 4.713684
details: Levant_N Han -0.000729 -2.884079
dscore: Han f4: 0.000018 Z: 0.145461

details: Nuer Karitiana -0.000506 -2.724728
details: Levant_N Karitiana 0.000383 0.972811
dscore: Karitiana f4: -0.000129 Z: -0.668439

details: Nuer Kostenki14 -0.001282 -4.102099
details: Levant_N Kostenki14 0.003253 4.936253
dscore: Kostenki14 f4: 0.000640 Z: 2.078251

details: Nuer MA1 -0.001476 -4.488804
details: Levant_N MA1 0.002483 3.934755
dscore: MA1 f4: 0.000202 Z: 0.649334

details: Nuer Mbuti 0.015175 53.235267
details: Levant_N Mbuti -0.020441 -40.335143
dscore: Mbuti f4: 0.000083 Z: 0.329452

details: Nuer Papuan 0.001172 5.028398
details: Levant_N Papuan -0.001842 -4.183511
dscore: Papuan f4: -0.000105 Z: -0.499191

details: Nuer Ust_Ishim 0.000647 2.152570
details: Levant_N Ust_Ishim -0.001157 -1.997243
dscore: Ust_Ishim f4: -0.000117 Z: -0.435765

gendstat: Chukchi Han 0.145
gendstat: Chukchi Karitiana -0.668
gendstat: Chukchi Kostenki14 2.078
gendstat: Chukchi MA1 0.649
gendstat: Chukchi Mbuti 0.329
gendstat: Chukchi Papuan -0.499
gendstat: Chukchi Ust_Ishim -0.436
gendstat: Han Karitiana -0.722
gendstat: Han Kostenki14 2.030
gendstat: Han MA1 0.579
gendstat: Han Mbuti 0.265
gendstat: Han Papuan -0.630
gendstat: Han Ust_Ishim -0.508
gendstat: Karitiana Kostenki14 2.323
gendstat: Karitiana MA1 0.972
gendstat: Karitiana Mbuti 0.725
gendstat: Karitiana Papuan 0.089
gendstat: Karitiana Ust_Ishim 0.038
gendstat: Kostenki14 MA1 -1.084
gendstat: Kostenki14 Mbuti -1.562
gendstat: Kostenki14 Papuan -2.319
gendstat: Kostenki14 Ust_Ishim -2.018
gendstat: MA1 Mbuti -0.316
gendstat: MA1 Papuan -0.879
gendstat: MA1 Ust_Ishim -0.834
gendstat: Mbuti Papuan -0.672
gendstat: Mbuti Ust_Ishim -0.637
gendstat: Papuan Ust_Ishim -0.040

## end of run



So, Somalis come out as:
62.6% Nuba
37.4% Levant_N


57.6% Nuer
42.4% Levant_N

with chisq 5.17 and 7.356 respectively.

blackflash16
08-28-2017, 05:27 AM
Copts:

Levant_N: 45%
Iran_ChL: 28%
Nuer: 27%

chisq 6.341 tail: 0.27

left pops:
Copt
Nuer
Levant_N
Iran_ChL

right pops:
Chukchi
Han
Karitiana
Kostenki14
MA1
Mbuti
Papuan
Ust_Ishim

0 Copt 19
1 Nuer 16
2 Levant_N 13
3 Iran_ChL 5
4 Chukchi 23
5 Han 43
6 Karitiana 12
7 Kostenki14 1
8 MA1 1
9 Mbuti 10
10 Papuan 14
11 Ust_Ishim 1
jackknife block size: 0.050
snps: 177419 indivs: 158
number of blocks for block jackknife: 705
## ncols: 177419
coverage: Copt 177344
coverage: Nuer 177323
coverage: Levant_N 134293
coverage: Iran_ChL 161249
coverage: Chukchi 177419
coverage: Han 177419
coverage: Karitiana 177419
coverage: Kostenki14 163879
coverage: MA1 126588
coverage: Mbuti 177419
coverage: Papuan 177419
coverage: Ust_Ishim 177094
dof (jackknife): 605.367
numsnps used: 88620
codimension 1
f4info:
f4rank: 2 dof: 5 chisq: 6.341 tail: 0.274426097 dofdiff: 7 chisqdiff: -6.341 taildiff: 1
B:
scale 1.000 1.000
Han 0.099 0.001
Karitiana -0.077 0.103
Kostenki14 -0.223 -1.030
MA1 -0.297 0.559
Mbuti 2.595 2.272
Papuan 0.247 -0.090
Ust_Ishim 0.232 -0.667
A:
scale 148.563 1324.832
Nuer 1.501 0.517
Levant_N -0.398 -1.088
Iran_ChL -0.767 1.244


full rank
f4info:
f4rank: 3 dof: 0 chisq: 0.000 tail: 1 dofdiff: 5 chisqdiff: 6.341 taildiff: 0.274426097
B:
scale 1.000 1.000 1.000
Han 0.096 0.080 -0.041
Karitiana -0.079 -0.005 0.139
Kostenki14 -0.195 -0.953 -0.883
MA1 -0.268 -0.513 1.056
Mbuti 2.598 2.345 2.128
Papuan 0.250 0.063 -0.144
Ust_Ishim 0.250 -0.563 -0.732
A:
scale 156.790 919.623 1671.909
Nuer 1.559 0.343 0.341
Levant_N -0.267 -1.578 -0.067
Iran_ChL -0.706 -0.627 1.697


best coefficients: 0.265 0.451 0.284
Jackknife mean: 0.265840692 0.438472001 0.295687307
std. errors: 0.012 0.167 0.157

error covariance (* 1000000)
144 -1699 1555
-1699 27902 -26203
1555 -26203 24649


summ: Copt 3 0.274426 0.266 0.438 0.296 144 -1699 1555 27902 -26203 24649

fixed pat wt dof chisq tail prob
000 0 5 6.341 0.274426 0.265 0.451 0.284
001 1 6 8.783 0.186137 0.247 0.753 -0.000
010 1 6 15.948 0.0140367 0.290 -0.000 0.710
100 1 6 16.752 0.0102387 0.000 4.834 -3.834 infeasible
011 2 7 6045.060 0 1.000 0.000 -0.000
101 2 7 482.178 0 0.000 1.000 -0.000
110 2 7 1024.974 0 0.000 -0.000 1.000
best pat: 000 0.274426 - -
best pat: 001 0.186137 chi(nested): 2.442 p-value for nested model: 0.118112
best pat: 101 5.42992e-100 chi(nested): 473.395 p-value for nested model: 5.86162e-105

coeffs: 0.265 0.451 0.284

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Nuer Han 0.000980 6.084157
details: Levant_N Han -0.000324 -1.390932
details: Iran_ChL Han -0.000498 -2.545316
dscore: Han f4: -0.000028 Z: -0.212677

details: Nuer Karitiana -0.000769 -3.319464
details: Levant_N Karitiana 0.000162 0.442643
details: Iran_ChL Karitiana 0.000529 1.749780
dscore: Karitiana f4: 0.000020 Z: 0.099526

details: Nuer Kostenki14 -0.002432 -5.661997
details: Levant_N Kostenki14 0.002103 3.388853
details: Iran_ChL Kostenki14 0.000555 1.011014
dscore: Kostenki14 f4: 0.000462 Z: 1.330290

details: Nuer MA1 -0.002618 -6.201165
details: Levant_N MA1 0.001283 2.143635
details: Iran_ChL MA1 0.002676 5.060852
dscore: MA1 f4: 0.000646 Z: 1.944139

details: Nuer Mbuti 0.027184 69.045117
details: Levant_N Mbuti -0.008563 -18.445620
details: Iran_ChL Mbuti -0.011141 -28.272082
dscore: Mbuti f4: 0.000173 Z: 0.630596

details: Nuer Papuan 0.002525 7.828418
details: Levant_N Papuan -0.000585 -1.426444
details: Iran_ChL Papuan -0.001252 -3.561636
dscore: Papuan f4: 0.000049 Z: 0.213026

details: Nuer Ust_Ishim 0.002176 5.321050
details: Levant_N Ust_Ishim 0.000485 0.899549
details: Iran_ChL Ust_Ishim -0.001461 -3.207086
dscore: Ust_Ishim f4: 0.000379 Z: 1.274037

gendstat: Chukchi Han -0.213
gendstat: Chukchi Karitiana 0.100
gendstat: Chukchi Kostenki14 1.330
gendstat: Chukchi MA1 1.944
gendstat: Chukchi Mbuti 0.631
gendstat: Chukchi Papuan 0.213
gendstat: Chukchi Ust_Ishim 1.274
gendstat: Han Karitiana 0.230
gendstat: Han Kostenki14 1.428
gendstat: Han MA1 2.052
gendstat: Han Mbuti 0.775
gendstat: Han Papuan 0.368
gendstat: Han Ust_Ishim 1.416
gendstat: Karitiana Kostenki14 1.184
gendstat: Karitiana MA1 1.797
gendstat: Karitiana Mbuti 0.481
gendstat: Karitiana Papuan 0.102
gendstat: Karitiana Ust_Ishim 1.058
gendstat: Kostenki14 MA1 0.437
gendstat: Kostenki14 Mbuti -0.752
gendstat: Kostenki14 Papuan -1.131
gendstat: Kostenki14 Ust_Ishim -0.209
gendstat: MA1 Mbuti -1.223
gendstat: MA1 Papuan -1.616
gendstat: MA1 Ust_Ishim -0.678
gendstat: Mbuti Papuan -0.435
gendstat: Mbuti Ust_Ishim 0.613
gendstat: Papuan Ust_Ishim 1.040




Nuba:

Nuer: 92%
Iran_ChL: 5%
Levant_N: 3%

chisq: 3.833 tail: 0.57


left pops:
Nuba
Nuer
Levant_N
Iran_ChL

right pops:
Chukchi
Han
Karitiana
Kostenki14
MA1
Mbuti
Papuan
Ust_Ishim

0 Nuba 16
1 Nuer 16
2 Levant_N 13
3 Iran_ChL 5
4 Chukchi 23
5 Han 43
6 Karitiana 12
7 Kostenki14 1
8 MA1 1
9 Mbuti 10
10 Papuan 14
11 Ust_Ishim 1
jackknife block size: 0.050
snps: 177419 indivs: 155
number of blocks for block jackknife: 705
## ncols: 177419
coverage: Nuba 177320
coverage: Nuer 177323
coverage: Levant_N 134293
coverage: Iran_ChL 161249
coverage: Chukchi 177419
coverage: Han 177419
coverage: Karitiana 177419
coverage: Kostenki14 163879
coverage: MA1 126588
coverage: Mbuti 177419
coverage: Papuan 177419
coverage: Ust_Ishim 177094
dof (jackknife): 604.782
numsnps used: 89465
codimension 1
f4info:
f4rank: 2 dof: 5 chisq: 3.833 tail: 0.573724383 dofdiff: 7 chisqdiff: -3.833 taildiff: 1
B:
scale 1.000 1.000
Han 0.104 0.050
Karitiana -0.092 0.045
Kostenki14 -0.159 -0.979
MA1 -0.376 0.083
Mbuti 2.581 2.388
Papuan 0.266 0.024
Ust_Ishim 0.290 -0.572
A:
scale 107.392 563.551
Nuer 0.104 0.078
Levant_N -1.058 -1.658
Iran_ChL -1.368 -0.495


full rank
f4info:
f4rank: 3 dof: 0 chisq: 0.000 tail: 1 dofdiff: 5 chisqdiff: 3.833 taildiff: 0.573724383
B:
scale 1.000 1.000 1.000
Han 0.109 0.075 0.117
Karitiana -0.084 -0.040 1.214
Kostenki14 0.002 -0.748 0.290
MA1 -0.414 -0.116 1.452
Mbuti 2.567 2.522 -0.191
Papuan 0.282 0.142 -0.633
Ust_Ishim 0.377 -0.204 -1.698
A:
scale 149.999 231.960 4285.708
Nuer 0.158 0.018 0.519
Levant_N -0.975 -1.439 0.661
Iran_ChL -1.423 -0.965 1.514


best coefficients: 0.923 0.029 0.048
Jackknife mean: 0.923928677 0.021336842 0.054734481
std. errors: 0.007 0.092 0.087

error covariance (* 1000000)
46 -493 447
-493 8439 -7946
447 -7946 7499


summ: Nuba 3 0.573724 0.924 0.021 0.055 46 -493 447 8439 -7946 7499

fixed pat wt dof chisq tail prob
000 0 5 3.833 0.573724 0.923 0.029 0.048
001 1 6 3.713 0.715424 0.920 0.080 -0.000
010 1 6 3.705 0.716556 0.925 -0.000 0.075
100 1 6 17.634 0.00721528 -0.000 15.668 -14.668 infeasible
011 2 7 287.680 0 1.000 -0.000 -0.000
101 2 7 3912.869 0 0.000 1.000 -0.000
110 2 7 5365.316 0 0.000 0.000 1.000
best pat: 000 0.573724 - -
best pat: 010 0.716556 chi(nested): -0.128 p-value for nested model: -nan
best pat: 011 2.5814e-58 chi(nested): 283.975 p-value for nested model: 1.02197e-63

coeffs: 0.923 0.029 0.048

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Nuer Han 0.000126 1.522209
details: Levant_N Han -0.001169 -4.085561
details: Iran_ChL Han -0.001339 -5.315042
dscore: Han f4: 0.000019 Z: 0.233460

details: Nuer Karitiana 0.000035 0.287303
details: Levant_N Karitiana 0.000957 2.181116
details: Iran_ChL Karitiana 0.001321 3.379920
dscore: Karitiana f4: 0.000123 Z: 1.023832

details: Nuer Kostenki14 -0.000001 -0.003716
details: Levant_N Kostenki14 0.004476 5.969861
details: Iran_ChL Kostenki14 0.002950 4.441269
dscore: Kostenki14 f4: 0.000270 Z: 1.299506

details: Nuer MA1 -0.000226 -1.045547
details: Levant_N MA1 0.003641 4.943023
details: Iran_ChL MA1 0.005020 7.204351
dscore: MA1 f4: 0.000137 Z: 0.655179

details: Nuer Mbuti 0.002893 15.700668
details: Levant_N Mbuti -0.032526 -53.773118
details: Iran_ChL Mbuti -0.035076 -64.457036
dscore: Mbuti f4: 0.000054 Z: 0.304320

details: Nuer Papuan 0.000230 1.540048
details: Levant_N Papuan -0.002856 -5.444831
details: Iran_ChL Papuan -0.003512 -7.532648
dscore: Papuan f4: -0.000038 Z: -0.262950

details: Nuer Ust_Ishim 0.000180 0.945187
details: Levant_N Ust_Ishim -0.001503 -2.176923
details: Iran_ChL Ust_Ishim -0.003424 -5.563528
dscore: Ust_Ishim f4: -0.000041 Z: -0.222431

gendstat: Chukchi Han 0.233
gendstat: Chukchi Karitiana 1.024
gendstat: Chukchi Kostenki14 1.300
gendstat: Chukchi MA1 0.655
gendstat: Chukchi Mbuti 0.304
gendstat: Chukchi Papuan -0.263
gendstat: Chukchi Ust_Ishim -0.222
gendstat: Han Karitiana 0.808
gendstat: Han Kostenki14 1.203
gendstat: Han MA1 0.575
gendstat: Han Mbuti 0.201
gendstat: Han Papuan -0.423
gendstat: Han Ust_Ishim -0.344
gendstat: Karitiana Kostenki14 0.660
gendstat: Karitiana MA1 0.060
gendstat: Karitiana Mbuti -0.344
gendstat: Karitiana Papuan -0.897
gendstat: Karitiana Ust_Ishim -0.808
gendstat: Kostenki14 MA1 -0.490
gendstat: Kostenki14 Mbuti -0.895
gendstat: Kostenki14 Papuan -1.386
gendstat: Kostenki14 Ust_Ishim -1.326
gendstat: MA1 Mbuti -0.332
gendstat: MA1 Papuan -0.754
gendstat: MA1 Ust_Ishim -0.719
gendstat: Mbuti Papuan -0.486
gendstat: Mbuti Ust_Ishim -0.466
gendstat: Papuan Ust_Ishim -0.017

## end of run


EDIT:
I tried to replicate the last qpAdm run from here (http://anthromadness.blogspot.ca/2016/07/somali-qpadm-models-using-new-ancient.html) using Nuer:

Somali:

Nuer: 58%
Levant_N: 31%
Iran_ChL: 12%


left pops:
Somali
Nuer
Levant_N
Iran_ChL

right pops:
Andamanese_Onge
Switzerland_HG
Chukchi
Han
Karitiana
Kostenki14
MA1
Mbuti.DG
Papuan
Ust_Ishim

0 Somali 13
1 Nuer 16
2 Levant_N 13
3 Iran_ChL 5
4 Andamanese_Onge 1
5 Switzerland_HG 1
6 Chukchi 23
7 Han 43
8 Karitiana 12
9 Kostenki14 1
10 MA1 1
11 Mbuti.DG 3
12 Papuan 14
13 Ust_Ishim 1
jackknife block size: 0.050
snps: 89518 indivs: 147
number of blocks for block jackknife: 702
## ncols: 89518
coverage: Somali 89518
coverage: Nuer 89470
coverage: Levant_N 67871
coverage: Iran_ChL 81357
coverage: Andamanese_Onge 89518
coverage: Switzerland_HG 89357
coverage: Chukchi 89518
coverage: Han 89518
coverage: Karitiana 89518
coverage: Kostenki14 82625
coverage: MA1 64010
coverage: Mbuti.DG 89517
coverage: Papuan 89518
coverage: Ust_Ishim 89363
dof (jackknife): 609.228
numsnps used: 44557
codimension 1
f4info:
f4rank: 2 dof: 7 chisq: 7.135 tail: 0.414932571 dofdiff: 9 chisqdiff: -7.135 taildiff: 1
B:
scale 1.000 1.000
Switzerland_HG 0.651 1.001
Chukchi 0.311 -0.085
Han 0.185 0.029
Karitiana 0.416 -0.216
Kostenki14 0.548 0.477
MA1 0.641 -0.841
Mbuti.DG -2.748 2.440
Papuan 0.000 0.495
Ust_Ishim 0.113 0.901
A:
scale 137.808 771.109
Nuer -0.815 -0.225
Levant_N 1.196 0.922
Iran_ChL 0.951 -1.449


full rank
f4info:
f4rank: 3 dof: 0 chisq: 0.000 tail: 1 dofdiff: 7 chisqdiff: 7.135 taildiff: 0.414932571
B:
scale 1.000 1.000 1.000
Switzerland_HG 0.647 0.965 -0.021
Chukchi 0.315 -0.046 0.818
Han 0.189 0.067 1.067
Karitiana 0.419 -0.178 -1.234
Kostenki14 0.549 0.510 1.134
MA1 0.643 -0.724 1.423
Mbuti.DG -2.747 2.501 0.376
Papuan 0.003 0.502 -0.359
Ust_Ishim 0.110 0.860 -1.445
A:
scale 135.842 702.375 5131.144
Nuer -0.795 -0.159 0.673
Levant_N 1.219 1.159 1.458
Iran_ChL 0.938 -1.278 0.648


best coefficients: 0.582 0.311 0.108
Jackknife mean: 0.582044125 0.296006770 0.121949105
std. errors: 0.008 0.103 0.100

error covariance (* 1000000)
69 -373 304
-373 10623 -10250
304 -10250 9945


summ: Somali 3 0.414933 0.582 0.296 0.122 69 -373 304 10623 -10250 9945

fixed pat wt dof chisq tail prob
000 0 7 7.135 0.414933 0.582 0.311 0.108
001 1 8 8.933 0.348014 0.579 0.421 -0.000
010 1 8 18.604 0.01713 0.595 -0.000 0.405
100 1 8 29.779 0.000231287 0.000 15.037 -14.037 infeasible
011 2 9 2092.615 0 1.000 0.000 -0.000
101 2 9 1082.946 0 0.000 1.000 -0.000
110 2 9 1708.631 0 0.000 0.000 1.000
best pat: 000 0.414933 - -
best pat: 001 0.348014 chi(nested): 1.797 p-value for nested model: 0.180036
best pat: 101 2.2183e-227 chi(nested): 1074.013 p-value for nested model: 1.47048e-235

coeffs: 0.582 0.311 0.108

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Nuer Switzerland_HG -0.003889 -6.743180
details: Levant_N Switzerland_HG 0.008257 6.559235
details: Iran_ChL Switzerland_HG 0.002845 2.596854
dscore: Switzerland_HG f4: 0.000610 Z: 1.221346

details: Nuer Chukchi -0.001639 -3.947720
details: Levant_N Chukchi 0.003617 4.285327
details: Iran_ChL Chukchi 0.002313 2.982070
dscore: Chukchi f4: 0.000419 Z: 1.196325

details: Nuer Han -0.000936 -2.406048
details: Levant_N Han 0.002657 3.400787
details: Iran_ChL Han 0.001284 1.764900
dscore: Han f4: 0.000420 Z: 1.286837

details: Nuer Karitiana -0.002401 -5.091934
details: Levant_N Karitiana 0.003961 4.165742
details: Iran_ChL Karitiana 0.003072 3.673814
dscore: Karitiana f4: 0.000165 Z: 0.409729

details: Nuer Kostenki14 -0.002934 -4.905378
details: Levant_N Kostenki14 0.007227 5.702538
details: Iran_ChL Kostenki14 0.003222 2.989143
dscore: Kostenki14 f4: 0.000886 Z: 1.732823

details: Nuer MA1 -0.003080 -5.129324
details: Levant_N MA1 0.006053 5.128220
details: Iran_ChL MA1 0.006044 5.677079
dscore: MA1 f4: 0.000740 Z: 1.495569

details: Nuer Mbuti.DG 0.015720 30.900223
details: Levant_N Mbuti.DG -0.019765 -19.592051
details: Iran_ChL Mbuti.DG -0.023547 -27.632840
dscore: Mbuti.DG f4: 0.000469 Z: 1.105871

details: Nuer Papuan -0.000046 -0.106502
details: Levant_N Papuan 0.001405 1.585994
details: Iran_ChL Papuan -0.000920 -1.197865
dscore: Papuan f4: 0.000311 Z: 0.847119

details: Nuer Ust_Ishim -0.000955 -1.742872
details: Levant_N Ust_Ishim 0.002514 2.422544
details: Iran_ChL Ust_Ishim -0.001083 -1.134583
dscore: Ust_Ishim f4: 0.000110 Z: 0.253338

gendstat: Andamanese_Onge Switzerland_HG 1.221
gendstat: Andamanese_Onge Chukchi 1.196
gendstat: Andamanese_Onge Han 1.287
gendstat: Andamanese_Onge Karitiana 0.410
gendstat: Andamanese_Onge Kostenki14 1.733
gendstat: Andamanese_Onge MA1 1.496
gendstat: Andamanese_Onge Mbuti.DG 1.106
gendstat: Andamanese_Onge Papuan 0.847
gendstat: Andamanese_Onge Ust_Ishim 0.253
gendstat: Switzerland_HG Chukchi -0.436
gendstat: Switzerland_HG Han -0.438
gendstat: Switzerland_HG Karitiana -0.947
gendstat: Switzerland_HG Kostenki14 0.524
gendstat: Switzerland_HG MA1 0.249
gendstat: Switzerland_HG Mbuti.DG -0.280
gendstat: Switzerland_HG Papuan -0.653
gendstat: Switzerland_HG Ust_Ishim -1.009
gendstat: Chukchi Han 0.002
gendstat: Chukchi Karitiana -0.966
gendstat: Chukchi Kostenki14 1.077
gendstat: Chukchi MA1 0.762
gendstat: Chukchi Mbuti.DG 0.139
gendstat: Chukchi Papuan -0.365
gendstat: Chukchi Ust_Ishim -0.876
gendstat: Han Karitiana -0.931
gendstat: Han Kostenki14 1.088
gendstat: Han MA1 0.752
gendstat: Han Mbuti.DG 0.144
gendstat: Han Papuan -0.407
gendstat: Han Ust_Ishim -0.894
gendstat: Karitiana Kostenki14 1.585
gendstat: Karitiana MA1 1.310
gendstat: Karitiana Mbuti.DG 0.739
gendstat: Karitiana Papuan 0.415
gendstat: Karitiana Ust_Ishim -0.135
gendstat: Kostenki14 MA1 -0.274
gendstat: Kostenki14 Mbuti.DG -0.860
gendstat: Kostenki14 Papuan -1.261
gendstat: Kostenki14 Ust_Ishim -1.558
gendstat: MA1 Mbuti.DG -0.536
gendstat: MA1 Papuan -0.953
gendstat: MA1 Ust_Ishim -1.276
gendstat: Mbuti.DG Papuan -0.418
gendstat: Mbuti.DG Ust_Ishim -0.836
gendstat: Papuan Ust_Ishim -0.525

## end of run

Awale
08-29-2017, 02:06 AM
Somali:

Nuer: 58%
Levant_N: 31%
Iran_ChL: 12%

The LNF + Zagrosian-Chalcolithic fitting well thing that David found really has me conflicted, to be honest. qpAdm makes it seem like it's good but every other analysis or line of evidence, as far as I know, indicates the contrary. ADMIXTURE usually just made Somalis seem predominantly Southwest Asian or more Southwest Asian + Mediterranean than notably part Caucasus alongside the prior two clusters. Then you have our mtDNA which is something I've pointed out to someone before:


Then we have Somalis' mtDNA N lineages like:

N1a
N1b
N1e
K1a
R0*
R0a
HV
U3a
U9a

These lineages mostly correlate with pre-historic samples from the Levant and Anatolia and not really with those from Iran:

Some Neolithic Anatolian mtDNA N lineages:

N1a
K1a
U3
N1b
J1
U8
W1
X2d
X2m
T2b
K1b
H5
H3a
H2a

Some Epipaleothic and Neolithic Levantine N lineages:

R0
R0a
K
K1a
U*
H
HV
N*
T1a

Some Neolithic Zagrosian N lineages:

R2
T2c
X2
J1c
J1d
K1a12a (a late Neolithic sample (6850Ī50ybp) that's also of Y-DNA G2a1)

Link for the Somali mtDNA results (http://anthromadness.blogspot.ae/2015/08/somali-mtdna-frequencies.html), forgive the condition of the charts, they used to be a lot better but google randomly changed their chart format on me and I'm thinking of using another outlet and embedding whole new charts soon. Link for the ancient mtDNA results (http://www.ancestraljourneys.org/wasianneolithicdna.shtml).

Basically, our mtDNA N lineages look overwhelmingly pre-Chalcolithic Anatolian-Levantine related. It doesn't look like the mtDNA profile of a heavily Bronze-Age Levantine influenced population which is what being 31% LNF + 12% Zagrosian-Chalcolithic implies our West-Eurasian side ought to be like. Nor do we display notable overlap with Neolithic Zagrosians, as you can see above. Even in tree-mix, if you use more "Basal-rich + Villabruna-related" populations like EEFs alongside ANEs like MA-1:

http://1.bp.blogspot.com/-mdanoevspbQ/VQJ6T38wz5I/AAAAAAAACRE/bEstjqcchBo/s1600/10p7uv6.jpg.png

We favor the more "Basal-rich + Villabruna-related" types entirely. And finally, there's nMonte based on global 10 PCA positions where the opposite of what goes on with qpAdm happens:

[1] "distance%=1.976 / distance=0.01976"

Somali

Dinka 57.4
Levant_EBA:I1706 42.6


Was reduced to using the Bronze Age Jordanians because the simulation won't even employ the use of Zagrosian-Chalcolithic alongside LNF.

-

[1] "distance%=1.2472 / distance=0.012472"

Somali

Dinka 57
Levant_N:I1704 43

-

[1] "distance%=0.6868 / distance=0.006868"

Somali

Dinka 55.9
Natufian:I0861 44.1

I dunno, think it says something when one analysis method is saying something that contradicts several others.

blackflash16
08-29-2017, 03:01 AM
The LNF + Zagrosian-Chalcolithic fitting well thing that David found really has me conflicted, to be honest. qpAdm makes it seem like it's good but every other analysis or line of evidence, as far as I know, indicates the contrary. ADMIXTURE usually just made Somalis seem predominantly Southwest Asian or more Southwest Asian + Mediterranean than notably part Caucasus alongside the prior two clusters. Then you have our mtDNA which is something I've pointed out to someone before:


Maybe we'll know more when the data/paper for that Tanzanian pastoralist is published. There's an interesting exchange in the comments here: (https://gnxp.nofe.me/2017/07/09/slc24a5-is-very-important-but-we-dont-know-why/)



Razib Khan: yes somalis share drift with the ancient iranian samples from the zagros. the ancient tanzanian individual does not share drift with the ancient iranian samples. it shares drift exclusively with the pre-pottery neolithic levant samples.


Also, what's with Mota. The d-stats show no significant drift with Eurasians compared with Yoruba, Dinka. Yet he seems to eat up Eurasian ancestry in Somalis compared to Nilotes:

Somalis:

Mota: 62%
Levant_N: 38%

chisq: 8.522 tail: 0.38


left pops:
Somali
Mota
Levant_N

right pops:
Andamanese_Onge
Switzerland_HG
Chukchi
Han
Karitiana
Kostenki14
MA1
Mbuti.DG
Papuan
Ust_Ishim

0 Somali 13
1 Mota 1
2 Levant_N 13
3 Andamanese_Onge 1
4 Switzerland_HG 1
5 Chukchi 23
6 Han 43
7 Karitiana 12
8 Kostenki14 1
9 MA1 1
10 Mbuti.DG 3
11 Papuan 14
12 Ust_Ishim 1
jackknife block size: 0.050
snps: 78761 indivs: 127
number of blocks for block jackknife: 701
## ncols: 78761
coverage: Somali 78761
coverage: Mota 78741
coverage: Levant_N 59894
coverage: Andamanese_Onge 78761
coverage: Switzerland_HG 78622
coverage: Chukchi 78761
coverage: Han 78761
coverage: Karitiana 78761
coverage: Kostenki14 72721
coverage: MA1 56358
coverage: Mbuti.DG 78760
coverage: Papuan 78761
coverage: Ust_Ishim 78625
dof (jackknife): 610.453
numsnps used: 41465
codimension 1
f4info:
f4rank: 1 dof: 8 chisq: 8.522 tail: 0.384183869 dofdiff: 10 chisqdiff: -8.522 taildiff: 1
B:
scale 1.000
Switzerland_HG 0.945
Chukchi 0.431
Han 0.323
Karitiana 0.471
Kostenki14 0.785
MA1 0.644
Mbuti.DG -2.547
Papuan 0.094
Ust_Ishim 0.261
A:
scale 153.387
Mota -0.747
Levant_N 1.201


full rank
f4info:
f4rank: 2 dof: 0 chisq: 0.000 tail: 1 dofdiff: 8 chisqdiff: 8.522 taildiff: 0.384183869
B:
scale 1.000 1.000
Switzerland_HG 0.957 0.374
Chukchi 0.435 -0.642
Han 0.324 -0.494
Karitiana 0.490 -0.558
Kostenki14 0.803 0.965
MA1 0.659 -0.161
Mbuti.DG -2.527 2.315
Papuan 0.106 0.878
Ust_Ishim 0.271 0.898
A:
scale 153.989 1353.168
Mota -0.645 1.295
Levant_N 1.259 0.569


best coefficients: 0.617 0.383
Jackknife mean: 0.616494915 0.383505085
std. errors: 0.014 0.014

error covariance (* 1000000)
209 -209
-209 209


summ: Somali 2 0.384184 0.616 0.384 209 -209 209

fixed pat wt dof chisq tail prob
00 0 8 8.522 0.384184 0.617 0.383
01 1 9 321.240 0 1.000 0.000
10 1 9 1054.419 0 0.000 1.000
best pat: 00 0.384184 - -
best pat: 01 8.08697e-64 chi(nested): 312.718 p-value for nested model: 5.58721e-70

coeffs: 0.617 0.383

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Mota Switzerland_HG -0.003222 -2.376654
details: Levant_N Switzerland_HG 0.008244 6.464715
dscore: Switzerland_HG f4: 0.001174 Z: 1.215343

details: Mota Chukchi -0.001988 -2.143923
details: Levant_N Chukchi 0.003505 4.117930
dscore: Chukchi f4: 0.000118 Z: 0.179990

details: Mota Han -0.001411 -1.617106
details: Levant_N Han 0.002620 3.298195
dscore: Han f4: 0.000135 Z: 0.222414

details: Mota Karitiana -0.002212 -2.130556
details: Levant_N Karitiana 0.003987 4.107710
dscore: Karitiana f4: 0.000164 Z: 0.221128

details: Mota Kostenki14 -0.001918 -1.438598
details: Levant_N Kostenki14 0.007205 5.757155
dscore: Kostenki14 f4: 0.001579 Z: 1.676519

details: Mota MA1 -0.002536 -1.898674
details: Levant_N MA1 0.005548 4.649444
dscore: MA1 f4: 0.000563 Z: 0.595169

details: Mota Mbuti.DG 0.013240 11.961357
details: Levant_N Mbuti.DG -0.019467 -19.256159
dscore: Mbuti.DG f4: 0.000701 Z: 0.923424

details: Mota Papuan 0.000680 0.717423
details: Levant_N Papuan 0.001396 1.581063
dscore: Papuan f4: 0.000954 Z: 1.441238

details: Mota Ust_Ishim 0.000181 0.157068
details: Levant_N Ust_Ishim 0.002792 2.675572
dscore: Ust_Ishim f4: 0.001182 Z: 1.449360

gendstat: Andamanese_Onge Switzerland_HG 1.215
gendstat: Andamanese_Onge Chukchi 0.180
gendstat: Andamanese_Onge Han 0.222
gendstat: Andamanese_Onge Karitiana 0.221
gendstat: Andamanese_Onge Kostenki14 1.677
gendstat: Andamanese_Onge MA1 0.595
gendstat: Andamanese_Onge Mbuti.DG 0.923
gendstat: Andamanese_Onge Papuan 1.441
gendstat: Andamanese_Onge Ust_Ishim 1.449
gendstat: Switzerland_HG Chukchi -1.257
gendstat: Switzerland_HG Han -1.263
gendstat: Switzerland_HG Karitiana -1.105
gendstat: Switzerland_HG Kostenki14 0.403
gendstat: Switzerland_HG MA1 -0.576
gendstat: Switzerland_HG Mbuti.DG -0.516
gendstat: Switzerland_HG Papuan -0.242
gendstat: Switzerland_HG Ust_Ishim 0.009
gendstat: Chukchi Han 0.055
gendstat: Chukchi Karitiana 0.100
gendstat: Chukchi Kostenki14 1.859
gendstat: Chukchi MA1 0.572
gendstat: Chukchi Mbuti.DG 0.946
gendstat: Chukchi Papuan 1.589
gendstat: Chukchi Ust_Ishim 1.686
gendstat: Han Karitiana 0.063
gendstat: Han Kostenki14 1.821
gendstat: Han MA1 0.542
gendstat: Han Mbuti.DG 0.933
gendstat: Han Papuan 1.724
gendstat: Han Ust_Ishim 1.665
gendstat: Karitiana Kostenki14 1.639
gendstat: Karitiana MA1 0.470
gendstat: Karitiana Mbuti.DG 0.737
gendstat: Karitiana Papuan 1.248
gendstat: Karitiana Ust_Ishim 1.339
gendstat: Kostenki14 MA1 -1.008
gendstat: Kostenki14 Mbuti.DG -1.021
gendstat: Kostenki14 Papuan -0.732
gendstat: Kostenki14 Ust_Ishim -0.454
gendstat: MA1 Mbuti.DG 0.152
gendstat: MA1 Papuan 0.459
gendstat: MA1 Ust_Ishim 0.672
gendstat: Mbuti.DG Papuan 0.369
gendstat: Mbuti.DG Ust_Ishim 0.649
gendstat: Papuan Ust_Ishim 0.321

beyoku
08-29-2017, 04:09 PM
Anyone up for extracting Y and Mitochondrial SNP's from the raw data?
It was sequenced, I have no idea how its done.

Lank
08-29-2017, 07:21 PM
These Beni-Amer look essentially Tigre (of course, it could also be the other way around) compared to the Hadendowa, perhaps not surprising considering the significant overlap between Beni-Amer and Tigre. It's nice to have another sample from the Horner "genetic continuum".

The K=18 looks like it picks up on recent Arabian ancestry in the Sudanese Arab groups rather well, peaking in the Bataheen, with a very low presence in the Nubian groups who live farther downstream on the Nile. The impact of Arabian migrants is fairly notable in Sudanese Arabs, although most of the Eurasian admixture looks more ancient.

Angoliga
08-29-2017, 11:40 PM
Interesting. Uniparentally, I know Nilo-Saharan types are dominated by y-DNA A and B and I believe maternally are predominantly L3. So do they get their West-African ancestry from their maternal side (I believe they also have some L2 mtDNA, correct me if I'm wrong).

Good question, I think the topic merits more discussion.

The answer depends on the "Nilo-Saharan type" in question but if we're to focus on "core" S-Sudanese populations, maybe something can be deduced en gros.

If one were to broadly stroke L2 as a marker for maternal West-African ancestry, and compare the mtDNA of populations like the Dinka, Nuer, Nuba and Shuilik, there does seems to be a proportionate correlation between their W-African affinity at K=4 (<20%) and their L2 mtDNA .

The rest of their mtDNA is mostly concentrated in East-Africa (L0,L3,L4,L5). (Mohammed, 2009 (http://ethiohelix.blogspot.ca/2015/06/sudan-mtdna.html))

*the Nuer might've been an exception in this tally due to their relatively smaller sample size (~less than half of most)






L2 mtDNA Distribution Map
*L2a would be the clear exception here*


Frequency distribution maps for mtDNA haplogroup L2. Maps for L2a (a), L2b (b), L2* (c), L2d (d), L2e (e) and total L2 (f).

http://i.imgur.com/aMc9yeR.png
(Silva, 2015 (https://www.researchgate.net/publication/280496425_60000_years_of_interactions_between_Cent ral_and_Eastern_Africa_documented_by_major_African _mitochondrial_haplogroup_L2))



From a plain point of view, my A-M13 marker and Central-Sudanic paternal grandmother's L2e is a clear demonstration (http://www.anthrogenica.com/showthread.php?241-L2d2-in-French-Guiana-and-beyond&p=147881&viewfull=1#post147881)of this.




Would anyone have references for the estimated dates/generations for the bulk of West-African admixture in these predominantly East-African populations?


Is it merely a distant sub-set of the East-African component?
Is there an explanation outside the great Bantu-Migrations? -- though the paper already established these NE populations were strongly endogamous and therefore unaffected
Could it partly be tied to returning migrations of East-Africans from West-Central Africa following the end of the Holocene Wet Phase?
This at least might be the case (http://www.anthrogenica.com/showthread.php?241-L2d2-in-French-Guiana-and-beyond&p=147881&viewfull=1#post147881) for Central Sudanics in East-Africa -- could it be admixture with these returning migrations?

Kaldo
08-31-2017, 11:26 AM
Would this confirm Copts (and by proxy, Ancient Egyptians) have some East African?

pgbk87
08-31-2017, 06:45 PM
Would this confirm Copts (and by proxy, Ancient Egyptians) have some East African?

Yes, it's safe to say ~10% East African.

Angoliga
11-26-2017, 07:32 PM
And the samples are available here (http://jakobssonlab.iob.uu.se/data/).

Would anyone know how to convert the sample data into txt (23andme) or csv (ftdna) files? I'd like to run them on some public admix calcs and share their results.

I've downloaded the gz file, extracted the files (BED, FAM, BIM) but seem stuck here... I'm using a windows OS, I've downloaded PLINK (dos 1.07) and tried their windows version (1.9) but both applications crash when I try to open them.


Thanks in advance for any help,


Edit: I've reached out to a few members on the forum but thought I'd also post directly on the thread

Angoliga
11-27-2017, 01:21 PM
Problem solved. I got some help from LukaszM, I was using cmd coding incorrectly. Here (http://www.anthrogenica.com/showthread.php?12744-K15-World-new-DIY-calculator&p=316364&viewfull=1#post316364)'s my first ever converted sample from a paper :nerd: -- I used a Zaghawa sample

I'll tabulate more the results in a spreadsheet as I go through this paper's dataset; I'm hoping to eventually convert the files for usage with GEDmatch calcs like puntDNAL

Angoliga
11-29-2017, 06:27 AM
If anyone's interested, the converted 23andme txt files for all 242 samples are here (https://www.dropbox.com/s/zh8sqdhvqugffc8/Sudan_Dataset_23%26me.7z?dl=0)

All samples have also now been uploaded onto GEDmatch, below are their kit#s:



Kit# ID

Z722611 Baria 93
Z018505 Baria_269
Z844808 Baria_271
Z926871 Baria_350
Z694433 Baria_412
Z585971 Bataheen_322
Z351889 Bataheen_323
Z825198 Bataheen_325
Z464159 Bataheen_327
Z303878 Bataheen_331
Z436503 Bataheen_332
Z948584 Bataheen_333
Z798967 Bataheen_334
Z616981 Bataheen_342
Z402337 Bataheen_465
Z455406 Bataheen_466
Z863128 Bataheen_467
Z596020 Bataheen_468
Z432182 Bataheen_470
Z057706 Bataheen_471
Z609667 Bataheen_474
Z361979 BeniAmer_189
Z357452 BeniAmer_190
Z369079 BeniAmer_195
Z132232 BeniAmer_200
Z634474 BeniAmer_202
Z029392 BeniAmer_203
Z749403 BeniAmer_207
Z906513 BeniAmer_208
Z069649 BeniAmer_219
Z628791 BeniAmer_230
Z263514 BeniAmer_239
Z733336 BeniAmer_243
Z476504 BeniAmer_245
Z082536 BeniAmer_246
Z906734 BeniAmer_251
Z851655 BeniAmer_255
Z184895 Copt_315
Z733385 Copt_316
Z717212 Copt_317
Z680607 Copt_318
Z624558 Copt_319
Z465026 Copt_441
Z970713 Copt_442
Z060371 Copt_443
Z622413 Copt_450
Z574086 Copt_451
Z574206 Copt_452
Z367864 Copt_453
Z043014 Copt_462
Z695448 Copt_463
Z207715 Copt_483
Z001842 Copt_485
Z057569 Danagla 185
Z163770 Danagla 64
Z939085 Danagla_115
Z030132 Danagla_120
Z548602 Danagla_121
Z941854 Danagla_122
Z736232 Danagla_133
Z912052 Danagla_140
Z387144 Danagla_144
Z312894 Danagla_145
Z090481 Danagla_153
Z282303 Danagla_154
Z587807 Danagla_155
Z193729 Danagla_157
Z767921 Danagla_159
Z166857 Danagla_173
Z725421 Dinka 75
Z022730 Dinka 90
Z447281 Dinka_101
Z907619 Dinka_267
Z983421 Dinka_272
Z338346 Dinka_305
Z895423 Dinka_391
Z391358 Dinka_397
Z947337 Dinka_403
Z381120 Dinka_405
Z826590 Dinka_413
Z512554 Dinka_424
Z067207 Dinka_448
Z398900 Dinka_481
Z513267 Dinka_60
Z040180 Dinka_61
Z873626 Gaalien_10
Z659474 Gaalien_112
Z229473 Gaalien_12
Z712114 Gaalien_15
Z313054 Gaalien_17
Z704487 Gaalien_20
Z917538 Gaalien_24
Z292419 Gaalien_28
Z348096 Gaalien_34
Z678434 Gaalien_4
Z302938 Gaalien_40
Z661789 Gaalien_47
Z131370 Gaalien_53
Z616879 Gaalien_54
Z793285 Gaalien_56
Z703639 Gemar 87
Z627603 Gemar 89
Z980699 Gemar 94
Z539240 Gemar 95
Z254458 Gemar 96
Z121676 Gemar 97
Z164046 Gemar_98
Z428182 Hadendowa_186
Z680417 Hadendowa_187
Z881248 Hadendowa_188
Z706411 Hadendowa_192
Z881516 Hadendowa_193
Z627041 Hadendowa_204
Z078934 Hadendowa_209
Z193352 Hadendowa_211
Z274750 Hadendowa_213
Z416640 Hadendowa_214
Z141842 Hadendowa_215
Z785347 Hadendowa_218
Z918637 Hadendowa_232
Z987212 Hadendowa_234
Z348187 Halfawieen_259
Z238671 Halfawieen_261
Z976419 Halfawieen_262
Z379123 Halfawieen_263
Z049601 Halfawieen_264
Z574456 Halfawieen_276
Z542470 Halfawieen_277
Z058169 Halfawieen_282
Z084488 Halfawieen_291
Z955978 Halfawieen_292
Z943332 Halfawieen_303
Z959172 Hausa_247
Z140677 Hausa_248
Z785686 Hausa_257
Z777881 Hausa_312
Z858536 Hausa_313
Z691838 Hausa_427
Z890326 Hausa_428
Z812026 Mahas_114
Z095474 Mahas_116
Z861477 Mahas_117
Z559707 Mahas_123
Z266132 Mahas_124
Z733405 Mahas_126
Z556586 Mahas_130
Z901310 Mahas_131
Z659510 Mahas_135
T420660 Mahas_136
T945709 Mahas_137
Z385771 Mahas_138
Z266731 Mahas_141
Z811655 Mahas_149
Z475566 Mahas_31
Z786939 Mahas_32
Z091619 Messiria 63
Z757212 Messiria 82
Z647357 Messiria_100
Z030051 Messiria_310
Z969201 Messiria_335
Z032925 Messiria_399
Z842399 Messiria_458
Z065143 Messiria_460
Z658469 Nuba 66
Z130173 Nuba 69
Z895188 Nuba 70
Z595323 Nuba 91
Z727557 Nuba 92
Z421340 Nuba_330
Z009481 Nuba_345
Z290510 Nuba_354
Z989321 Nuba_385
Z695686 Nuba_389
Z482434 Nuba_390
Z007102 Nuba_398
Z009470 Nuba_409
Z620862 Nuba_414
Z465667 Nuba_415
Z849014 Nuba_62
Z616206 Nuer 482
Z702611 Nuer_268
Z577509 Nuer_420
Z705919 Nuer_421
Z104416 Nuer_422
Z727597 Nuer_423
Z447626 Nuer_429
Z445940 Nuer_449
Z923886 Nuer_454
Z734435 Nuer_455
Z320251 Nuer_456
Z747660 Nuer_461
Z124038 Nuer_495
Z892295 Nuer_496
Z494516 Nuer_497
Z644679 Nuer_59
Z079015 Shaigia 77
Z872549 Shaigia_273
Z744349 Shaigia_304
Z551982 Shaigia_307
Z807699 Shaigia_308
Z032718 Shaigia_326
Z790789 Shaigia_328
Z167345 Shaigia_329
Z571354 Shaigia_382
Z924421 Shaigia_384
Z027928 Shaigia_401
Z143597 Shaigia_417
Z688248 Shaigia_418
Z790513 Shaigia_419
Z089915 Shaigia_446
Z907603 Shilluk_265
Z514155 Shilluk_266
Z250865 Shilluk_270
Z031606 Shilluk_274
Z395708 Shilluk_284
Z662306 Shilluk_285
Z822321 Shilluk_286
Z444917 Shilluk_287
Z025428 Shilluk_289
Z424386 Shilluk_392
Z406867 Shilluk_393
Z436390 Shilluk_394
Z266331 Shilluk_395
Z593969 Shilluk_396
Z253250 Shilluk_407
Z371131 Shilluk_480
Z641464 Zagawa 435
Z174487 Zagawa 67
Z553102 Zagawa_103
Z739955 Zagawa_356
Z941370 Zagawa_358
Z495833 Zagawa_361
Z775915 Zagawa_365
Z405386 Zagawa_368
Z086893 Zagawa_369
Z403388 Zagawa_373
Z689979 Zagawa_378
Z498099 Zagawa_379
Z283937 Zagawa_381
T005064 Zagawa_436
T745992 Zagawa_437
Z169251 Zagawa_80



Summary

Ethnicity, Qty:

Baria, 5
Bataheen, 16
BeniAmer, 16
Copt, 16
Danagla, 16
Dinka, 16
Gaalien, 15
Gemar, 7
Hadendowa, 14
Halfawieen, 11
Hausa, 7
Mahas, 16
Messiria, 8
Nuba, 16
Nuer, 16
Shaigia, 15
Shilluk, 16
Zagawa, 16



*Ethnicity/Sample Qty Summary is at the bottom of the code

Awale
11-29-2017, 08:35 AM
Problem solved. I got some help from LukaszM, I was using cmd coding incorrectly. Here (http://www.anthrogenica.com/showthread.php?12744-K15-World-new-DIY-calculator&p=316364&viewfull=1#post316364)'s my first ever converted sample from a paper :nerd: -- I used a Zaghawa sample

I'll tabulate more the results in a spreadsheet as I go through this paper's dataset; I'm hoping to eventually convert the files for usage with GEDmatch calcs like puntDNAL

Yeah, man. Sorry, I don't have much experience with this sort of thing (just did a little reading on how it all works with King a while back) and I've been busy with offline stuff lately. Good work getting the samples you so far have, though. I must admit that I lost interest in this study's samples after David told me there wasn't enough SNP overlap for him to throw them into his global-PCA (dunno if he could run some formal-stats on them, though).

puntDNALKing
11-29-2017, 11:26 AM
I ran few people through my Global22DNA puntDNAL K15 and here are the results:

Disclaimer: I only ran a single person for each population.


According to my results (which don't add-up to 100% because I only used the necessary components), I found that the Beja (BeniAmer) had increased Nilotic component (as expected from a populations in Sudan) and somewhat increase omotic component when compared to the Arab Sudanese population. This omotic component is also found in Ethiopia and Somali. A West African component is found among Sudanese Arabs, but is absent among the Beja and Horners. An increased Mediterranean component is found among the Sudanese Arabs and not among the Beja and Horners. An increased Caucasian Component is found among the Sudanese and to some degree among the Northern Ethopians, but is almost absent in Somali.

This is just a preliminary run; tell me what you guys think.

Angoliga
11-30-2017, 05:25 AM
All 242 samples have now been uploaded onto GEDmatch, their kit#s are now posted in the previous (http://www.anthrogenica.com/showthread.php?11784-New-High-Coverage-Northeast-African-genomes&p=317347&viewfull=1#post317347) post




I ran few people through my Global22DNA puntDNAL K15 and here are the results:

Disclaimer: I only ran a single person for each population.


According to my results (which don't add-up to 100% because I only used the necessary components), I found that the Beja (BeniAmer) had increased Nilotic component (as expected from a populations in Sudan) and somewhat increase omotic component when compared to the Arab Sudanese population. This omotic component is also found in Ethiopia and Somali. A West African component is found among Sudanese Arabs, but is absent among the Beja and Horners. An increased Mediterranean component is found among the Sudanese Arabs and not among the Beja and Horners. An increased Caucasian Component is found among the Sudanese and to some degree among the Northern Ethopians, but is almost absent in Somali.

This is just a preliminary run; tell me what you guys think.

Nice! Looking forward to see what happens with the other pops -- from a first glance these observations seem inline with what one would've expected.


Observations:


It's perplexing that these runs significantly show the very ancient Nilotic-like affinity in the Somali (~40%) yet the ancient West-African affinity in the Nuba is only slightly apparent (~10%).
Isn't the ancient West-African for the Nuba typically nearly half their ancestry in the lower ks? I don't think they're as pristinely "Nilotic" as S.Sud pops like the Dinka, Nuer, Gumuz & co


By any chance, would you have a North-African component in the works for this new calc'? Perhaps something similar to what we've seen peaking in Copts. It'd be neat seeing this component contrasting against a more SW-Asian affinity that peaks in Yemeni/Bedouin pops -- it'd be really interesting seeing how these ratios appear among Northern-Sudanese/Horners.


Last but not least, now that we have the samples, would there be difficulty teasing a 2nd "Nilotic" component? I have several ADMIX references of these components in recent papers if you'd like.
Past papers have called this component Darfurian/Chadic/Central-Sudanc, I like the term Sahelian. In this paper (Hollfelder, 2017) it peaks among Darfurians like the Zagawa and the "Gemar". If this cluster were included, it would definitely make for a better oracle distinguishing Central-Sudanic/Chadic/Darfurian speakers vs. the Nilotic pops which dominate Southern-Sudan (Dinka/Nuer).

Batroun
11-30-2017, 05:38 AM
off Topic: Ive tried reading through this thread but given that im only starting to understand how all of this stuff works, where can I find more info on Copts and compare them to Ancient Egyptians and Muslim Egyptians

puntDNALKing
11-30-2017, 08:49 AM
It's perplexing that these runs significantly show the very ancient Nilotic-like affinity in the Somali (~40%) yet the ancient West-African affinity in the Nuba is only slightly apparent (~10%).
Isn't the ancient West-African for the Nuba typically nearly half their ancestry in the lower ks? I don't think they're as pristinely "Nilotic" as S.Sud pops like the Dinka, Nuer, Gumuz & co

According to my previous runs of Nuba, they score 12-18% West African never half (if I recall correctly). So you can think of Nuba and Gumuz as two sides of the coin, nuba with minor West African affinity and Gumuz with minor Omotic affinity. Usually, Gumuz score 65-75% Nilotic with the rest being Omotic and Nuba score 75-80% Nilotic with the rest being West African.


By any chance, would you have a North-African component in the works for this new calc'? Perhaps something similar to what we've seen peaking in Copts. It'd be neat seeing this component contrasting against a more SW-Asian affinity that peaks in Yemeni/Bedouin pops -- it'd be really interesting seeing how these ratios appear among Northern-Sudanese/Horners.


Yes, I will have a North African component. I hope it is a NE African component as opposed to a NW African component; I will know by tomorrow.


Last but not least, now that we have the samples, would there be difficulty teasing a 2nd "Nilotic" component? I have several ADMIX references of these components in recent papers if you'd like.
Past papers have called this component Darfurian/Chadic/Central-Sudanc, I like the term Sahelian. In this paper (Hollfelder, 2017) it peaks among Darfurians like the Zagawa and the "Gemar". If this cluster were included, it would definitely make for a better oracle distinguishing Central-Sudanic/Chadic/Darfurian speakers vs. the Nilotic pops which dominate Southern-Sudan (Dinka/Nuer).

Yes, I will try to run a african only calculator with this data in the coming month or so. Hopefully, this will give us a better breakdown of the North African and Nilotic components.

Awale
11-30-2017, 08:57 AM
Yes, I will have a North African component. I hope it is a NE African component as opposed to a NW African component; I will know by tomorrow.

I'll email you the Copt samples I have, should be able to help you get a NE-African cluster using them, I hope.

puntDNALKing
11-30-2017, 09:12 AM
I'll email you the Copt samples I have, should be able to help you get a NE-African cluster using them, I hope.

Thanks a lot, I only have 2 Copts samples in my private collection plus the 16 or so in this dataset.

puntDNALKing
12-04-2017, 01:38 AM
I am currently running the Global22DNA admixture again, with the added copt dna. Here is a global MDS C1 v C2:
20223

Here is a close up on the Africa data:

HOA and Sudan:
20224

Nilotic/ North Bantu:
20225

West Africa:
20226


HG South Africa/ Central Africa:
20227

puntDNALKing
12-04-2017, 01:42 AM
Continued...

North Africa:
20232

Omotic/ north bantu:
2023120231

puntDNALKing
12-05-2017, 03:32 AM
20251

Here is my old pca with just Africa. This one gives better indication of the gene spread across Africa.

Lank
12-05-2017, 09:03 PM
Someone uploaded the samples from this study to GEDmatch, so I compared the Beni Amer and Hadendowa. Interestingly, there is very little IBD matching between these groups. The Beni Amer tend to match each other, and unlike the Hadendowa, they often match lowland Eritreans (Tigre), and occasionally also highlander Eritreans/Ethiopians (I match one). The Hadendowa on the other hand also match each other, but then they also tend to match the academic North Sudanese Arab samples, which fits with their more Sudanese-like ADMIXTURE results.

Angoliga
12-06-2017, 12:41 AM
Someone uploaded the samples from this study to GEDmatch, so I compared the Beni Amer and Hadendowa. Interestingly, there is very little IBD matching between these groups. The Beni Amer tend to match each other, and unlike the Hadendowa, they often match lowland Eritreans (Tigre), and occasionally also highlander Eritreans/Ethiopians (I match one). The Hadendowa on the other hand also match each other, but then they also tend to match the academic North Sudanese Arab samples, which fits with their more Sudanese-like ADMIXTURE results.

Nice! you had a match as well?! I was hoping to observe a correlation between which side of the family matched particular ethnicities but couldn't find any. Both Central and Eastern Sudanic speakers of my family matched a handful of upper-nilotes (Nuer, Baria, Shilluk).
I skimmed through a few kits last week and found it interesting that some Messeria didn't have any matches at all whatsoever, not even among themselves down to 5cM segments!

Considering how endogamous Copts are proven to be, I was curious to see if any of our Coptic anthrogenica members came across any new matches (need4speed?) :)



.. apropos, I uploaded the kits last week -- it took me 2 evenings to convert into 23andme txt files and manually upload all 242 samples (http://www.anthrogenica.com/showthread.php?11784-New-High-Coverage-Northeast-African-genomes&p=317347&viewfull=1#post317347) :biggrin1:

I was initially stuck on how to convert the sample data in Plink but got some pointers from LanksM and Khanabadoshi. When it came to creating a new GEDmatch account, I couldn't think of a cool name so just stuck with something generic; "genedataset" - @gmail.com



EDIT: In other news, it looks like we finally have a Sahelian component in the works (http://www.anthrogenica.com/showthread.php?12788-New-K30-World-calculator&p=321080&viewfull=1#post321080):nerd:

Awale
12-07-2017, 12:13 AM
The Beni Amer tend to match each other, and unlike the Hadendowa, they often match lowland Eritreans (Tigre)

This in particular makes a lot of sense since Tigres' language literally has a Beja/North-Erythraeic substrate (http://anthromadness.blogspot.com/2017/08/being-more-accurate-about-substrates-in.html) and they're also semi-nomadic pastoralists. Plus, many Tigres are actually considered part of the Beni-Amer tribe, from what I recall. I'd love to see how these guys relate to nearby Sahos and Afars someday, however. I think there's long been a cultural exchange route along the NE African coast along the lines of Beja->Saho->Afar->Somali and possibly backwards due to the long time similarity in material culture and some customs between these ethnic groups, it'd be interesting to see if some of this translates into gene flow.

SWAHILLI_PRINCE16
12-07-2017, 12:40 AM
Looks like I'm a distant relative of Nilosaharan :)

GEDmatch Genesis
Nilosaharan
7.269
7.269
7.5
302778

Angoliga
12-07-2017, 02:58 AM
Looks like I'm a distant relative of Nilosaharan :)

GEDmatch Genesis
Nilosaharan
7.269
7.269
7.5
302778

Hmm, very peculiar... it looks like you've matched one of the samples I first uploaded onto GEDmatch Genesis -- the kit isn't mine, it belongs to a Nuer (https://en.wikipedia.org/wiki/Nuer_people)sample.
When first trying to convert the 242 samples from the paper, I wasn't able to convert into 23andMe txt format (11 Mb file size); I could only convert into huge VCF files (111Mb) which are only uploadable on Genesis. After a few extremely long uploads, I reached out for help on how to convert into smaller txt format.
So, I think the higher resolution in those VCF genomes might've allowed for more matches because that same sample ("Nuer_420", #Z577509) has 0 matches on regular GEDmatch whereas on Genesis the kit somehow has 755 matches (#SY9579367)!! Even more peculiar, all these 755 matches are quite sizable, they range from 7-17 Total cM :confused:

It's a similar scenario with the other few VCF samples uploaded on Genesis:


#PV8212613 Zagawa 381
#HU6566248 Nuba62
#RN9482968 Nuba330
#HM5967703 Nuer268


Perhaps a more knowledgeable member on the forum could speak to this?

@SwahiliPrince don't fret though, we're still fam :)
If I recall correctly, your Dad's kit is a distant GEDmatch to one of my paternal aunts -- albeit <8cM ;)

Lank
12-07-2017, 03:44 AM
This in particular makes a lot of sense since Tigres' language literally has a Beja/North-Erythraeic substrate (http://anthromadness.blogspot.com/2017/08/being-more-accurate-about-substrates-in.html) and they're also semi-nomadic pastoralists. Plus, many Tigres are actually considered part of the Beni-Amer tribe, from what I recall. I'd love to see how these guys relate to nearby Sahos and Afars someday, however. I think there's long been a cultural exchange route along the NE African coast along the lines of Beja->Saho->Afar->Somali and possibly backwards due to the long time similarity in material culture and some customs between these ethnic groups, it'd be interesting to see if some of this translates into gene flow.
Yeah, and a lot of Beni Amer actually speak Tigre. They may not actually be distinguishable from Tigres at all based on these samples, although there may be some variation we're missing. It would be interesting to see if members of the nabtab ruling caste are distinguishable or more related to the other Sudanese populations.

By the way, there is a lot of Y-DNA T among the GEDmatch users (coastal northeastern Sudanese and a Tigre from Massawa) matching these samples.

puntDNALKing
12-08-2017, 02:50 AM
This in particular makes a lot of sense since Tigres' language literally has a Beja/North-Erythraeic substrate (http://anthromadness.blogspot.com/2017/08/being-more-accurate-about-substrates-in.html) and they're also semi-nomadic pastoralists. Plus, many Tigres are actually considered part of the Beni-Amer tribe, from what I recall. I'd love to see how these guys relate to nearby Sahos and Afars someday, however. I think there's long been a cultural exchange route along the NE African coast along the lines of Beja->Saho->Afar->Somali and possibly backwards due to the long time similarity in material culture and some customs between these ethnic groups, it'd be interesting to see if some of this translates into gene flow.

I agree, my global22DNA PCA shows an affinity between highland ethiopians and the beja.

20304
20305

Angoliga
12-10-2017, 09:13 PM
Łukasz just released his K45 admix calc (http://www.anthrogenica.com/showthread.php?12788-New-K30-World-calculator&p=323052&viewfull=1#post323052) with the recently proposed Sahelian, Omotic and E_HG components

I ran some samples from the paper to cover the major groups (Nubian, Beja, Nilote, Copt, Hausa, Arab). Their results have been posted on his thread here (http://www.anthrogenica.com/showthread.php?12788-New-K30-World-calculator&p=323302&viewfull=1#post323302) in a spreadsheet along with some observations

nee4speed111
01-22-2018, 06:11 AM
Nice! you had a match as well?! I was hoping to observe a correlation between which side of the family matched particular ethnicities but couldn't find any. Both Central and Eastern Sudanic speakers of my family matched a handful of upper-nilotes (Nuer, Baria, Shilluk).
I skimmed through a few kits last week and found it interesting that some Messeria didn't have any matches at all whatsoever, not even among themselves down to 5cM segments!

Considering how endogamous Copts are proven to be, I was curious to see if any of our Coptic anthrogenica members came across any new matches (need4speed?) :)



.. apropos, I uploaded the kits last week -- it took me 2 evenings to convert into 23andme txt files and manually upload all 242 samples (http://www.anthrogenica.com/showthread.php?11784-New-High-Coverage-Northeast-African-genomes&p=317347&viewfull=1#post317347) :biggrin1:

I was initially stuck on how to convert the sample data in Plink but got some pointers from LanksM and Khanabadoshi. When it came to creating a new GEDmatch account, I couldn't think of a cool name so just stuck with something generic; "genedataset" - @gmail.com



EDIT: In other news, it looks like we finally have a Sahelian component in the works (http://www.anthrogenica.com/showthread.php?12788-New-K30-World-calculator&p=321080&viewfull=1#post321080):nerd:

Hey sorry for the late reply, some of the Sudanese copts that you sampled did in fact match with some of the Egyptian coptic samples I already have, which is pretty interesting considering that they have probably been part of seperate gene pools for close to 2 centuries now.

Angoliga
05-31-2018, 11:36 PM
Diversity in Sudan: Here's an interesting clip showcasing some of the ethnic groups mentioned in this paper


https://www.youtube.com/watch?v=kZLJ1CRPUjI

ThaYamamoto
01-30-2021, 04:30 AM
Anyone know why the SNP overlap for these was too low for g25? Is the data unphased and would the case be different now ?

Michalis Moriopoulos
10-31-2021, 04:12 AM
Anyone know why the SNP overlap for these was too low for g25? Is the data unphased and would the case be different now ?

This should definitely be revisited soon, especially with the coming publication of Kulubnarti which will apprently feature those Hollfelder genomes in some capacity (at least based on the preprint).