Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
ADMIXTURE @ K=11 limited only to Afr, SWAsia, NEur to compare and contrast gainst Schuenemann's Natufian Brown.
aar8380_vandeLoosdrecht_SM Loosdrecht2018 Fig. S11 redux from south(Ju Hoan) to north then from west(Sahrawi) to east, basically. Had to insert Natufian and Levant Neolithic twice.
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
This thread is more or less to show different researchers have different ADMIXTURE results about Natufian components and elements.
quote: From Natufian Brown OP:
I like to see several K's when I go over-interpreting STRUCTURE. At level 16 global there's a lot of continental substructure. Plus without lower K's I can't track progression of colors in populations as K increases. That's important as will be seen examining Natufians in other researchers STRUCTURE graphs. Colors can change frequency, pop in, and drop out of any population.
Here's Loosdrecht2018 Natufian K-volution
*01 GREEN maximized in Neolithic Levant _02 PURPLE 100% in Atlantic to Bight of Benin *03 ORANGE minor appearance in Hadza, Mota, Luxmanda and Taforalt. *05 RED minor appearance in Hadza, Mota, Luxmanda, and Taforalt. *07 MINT minor appearance in Dinka, Oromo, Afar, Somali, and Taforalt. *10 BROWN 100% in Taforalt. *11 GREY 100% in Hadza.
* indicates K=11 components and elements (no West African)
Here Natufians are 79% southwest Asian, 14% Taforalt, 6% East&Horn, and 1% Hadza "genetic components".
It's not for me to make a graph analysis conform to a priori. I interpret the data for what it is. No one can say now I'm calling Natufian 4/5ths 'Eurasian', it's Loos' data. Multidisciplinary vetting decides if impossible, possible, or probable. How I'd like it to be is not a criteria.
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
Taforalt K-volution
Showing pre-LGAM components and elements. Note K=9's 6-way substructure.
01 GREEN maximized in Neolithic Levant 02 PURPLE 100% in Atlantic to Bight of Benin 03 ORANGE minor appearance in Hadza, Mota, Luxmanda, and Natufian. 04 INDIGO 100% in Ju Hoan. 05 RED minor appearance in Hadza, Mota, Luxmanda, and Natufian. 07 MINT minor appearance in Dinka, Oromo, Afar, Somali, and Natufian.
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
Loosedrecht & Bouzouggar threads months ago relied on K=10. Here's 2 9 10 11 and redux.
Check what the authors say about their main text chosen K=10 ADMIXTURE.
Look what happens to Hadza and Taforalt when another bin opens after K=10. Note the Rainforest and EAfr (Luxmanda Tanzania & Mota to Somalia & Horn + central Sudan) changes and temporal continuity.
It looks like they are stating that McEvoy et al was correct.
Paleolithic, Holocene there was always been continuous movements.
question - are you running these analysis using your own PC using downloaded genome files?
-------------------- Without data you are just another person with an opinion - Deming Posts: 12143 | From: When you have eliminated the impossible, whatever remains, however improbable | Registered: Jun 2007
| IP: Logged |
posted
So this "uncharacterized" DNA means that this population no longer exist but modern West Africans carry "remnants" of this ghost population. Maybe someone can include the Melanesians/Papuans in the mix. All indications are Iwo-Eleru may be related to Papuans and Australians. Keeping in mind the Taforalt are also mixed with this "ghost" sub Saharan African population. Also Melanesians show close genetic affinity to African in certain Cluster charts while in others they are very distant. So it depends on what the targeted SNPS are in the cluster charts. "manipulation" can throw interpretation off and skew the inferences.
Scientists discover genomic ancestry of Stone Age North Africans from Morocco Ancient nuclear DNA from 15,000-year-old modern humans from Morocco, the oldest ever recovered from Africa, shows dual genetic ancestry to ancient Near Eastern and to sub-Saharan African populations Date: March 15, 2018 Source: Max Planck Institute for the Science of Human History Summary: An international team of researchers have sequenced DNA from individuals from Morocco dating to approximately 15,000 years ago. This is the oldest nuclear DNA from Africa ever successfully analyzed. The study shows that the individuals, dating to the Late Stone Age, had a genetic heritage that was in part similar to ancient Levantine Natufians and an uncharacterized sub-Saharan African lineage to which modern West Africans are genetically closest.
-------------------- Without data you are just another person with an opinion - Deming Posts: 12143 | From: When you have eliminated the impossible, whatever remains, however improbable | Registered: Jun 2007
| IP: Logged |
posted
As I said many times before...."Arabia is an extension of Africa". It is very easy to see what went down in pre-history. North Africa and Levant was one geographic area occupied by the same people(not modern Turks). SSA Extended from West Africa to Yemen. BOTH(Arabia/Africa) regions are occupied by Africans. That is why Arabia has such a high frequency of mtDNA L which extends to Harrapan Valley and off course Persia. E1b1b-M35 is older than E-M2 that is why the frequency of E-M2 is lower outside Africa compared to E-M35. All these lineages have a SSA origin in Southern East Africa. --- Quote" " The high proportion of Near Eastern ancestry shows that the connection between North Africa and the Near East began much earlier than many previously thought. Although the connections between these regions have been shown in previous studies for more recent time periods, it was not generally believed that humans were interacting across these distances during the Stone Age. "Our analysis shows that North Africa and the Near East, even at this early time, were part of **one region **without much of a genetic barrier," explains co-senior author Choongwon Jeong."
---
So which NA population has the highest amount of SSA ancestry? Yep. MAzb and Tunisians. So are these two groups closest to Taforalt? Forget the visuals!!
---
Quote: "Although the Sahara did present a physical barrier, there was also clearly interaction happening at this time. The strong connection between the Taforalt individuals and sub-Saharan populations shows that interactions across this vast desert were occurring much earlier than was previously thought. In fact, the proportion of sub-Saharan ancestry of the Taforalt individuals, one-third, is a higher percentage than found in modern populations in Morocco and many other North African populations.
Sub-Saharan heritage from a previously unknown ancient population Though the scientists found clear markers linking the heritage in question to sub-Saharan Africa, no previously identified population has the precise combination of genetic markers that the Taforalt individuals had. While some aspects match modern Hadza hunter-gatherers from East Africa and others match modern West Africans, **neither of these groups **has the same combination of characteristics as the Taforalt individuals. Consequently, the researchers cannot be sure exactly where this heritage comes from. One possibility is that this heritage may come from a population that no longer exists. However, this question would need further investigation."
-------------------- Without data you are just another person with an opinion - Deming Posts: 12143 | From: When you have eliminated the impossible, whatever remains, however improbable | Registered: Jun 2007
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
quote:Originally posted by xyyman: question - are you running these analysis using your own PC using downloaded genome files?
No. When I do a redux I don't run anything. I use the genuine ADMIXTURE graph just as the author(s) released it. I simply rotate the bars, so page doesn't get too wide, and reorder them. Pending focus I consider any combination of majority/plurality component(s) geography language trying my best to produce smooth clines.
I don't own the resources, don't know where to rent them, and refuse to spend rum cigar music & women $$$ on it.
The analysis part is just reporting exactly what I observe. I start by zooming out and looking at the color patterns. Then I comment on what I see about region, group, and individuals.
Keep in mind geneticists/statisticians make several runs of all K levels. They publish the one in most agreement with the others. Say out of 10 runs 2 are squirrely, 1 is off the wall, 7 are reasonable. Of the 7 say 3 strongly agree. They choose 1 of them. They further choose the most reliable K of all K levels in the chosen 1.
Note that Loosdrecht's Natufian components & frequencies vary compared to Schuenemann's Natufian result. Unfortunately Schuenemann did not publish any other than K=16. The two can still compared though I'd like to do so using the same K level for both.
Regardless of level Loosdrecht and Schuenemann components & elements and their frequencies cannot be the same. Schue's more robust, including the Americas, Oceania, and Asia. That will influence the results. For instance, is the Onge maximized element in one of her Natufians really repping residual OoA matter?
But when an "independent" source (like DNATribes) can get hold and manipulate the dataset(s) we will get some real answers. But already based upon what was selectively released we are getting answers.
So far it looks like only ELmaestro has the tools and knowledge to do independent runs. But he is playing the cards close to his chest
-------------------- Without data you are just another person with an opinion - Deming Posts: 12143 | From: When you have eliminated the impossible, whatever remains, however improbable | Registered: Jun 2007
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
EM's a busy guy.
ES is bound to get a member with the resources and time to field tests we might design. Try to stay alive until then!
Sample populations mean everything. Loosdrecht demonstrated how the dice get loaded. See her Fig. S12 outgroup-f3 effect. Then her following figures showing drastically different results when fellow African Mbuti are the outgroup vs distant non-African UshIshtim.
UBBFriend: Email this page to someone!
Printer-friendly view of this topic