posted
So I take it that you comitting to the idea that the 23% Khoisan-like component is the reason why hadza are farther from north East Africans than yorubans even though when that is taken out they(the hadza) assumes a position closest to where you beleive north East Africans are? Let’s get this on record.
& I don’t care about your judge of character.. I don’t care about your constant warping and distortion of my words anymore. I don’t even care about your insults... I just wanna hear about these Iranians being more ancient north East African than all population/samples seen thus far in and out of Africa... But also somehow closest to yorubans (fst) out of all unadmixed Africans.
That’s all I’ll take into consideration from here on out...
@tukuler... did you try to post an image? Fula 1 and two are both supposedly Fulani, they’re only separated because admixture picked up substructure within the ethnic group. The division is aesthetic.
Posts: 1782 | From: New York | Registered: Jul 2016
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
I tried to find out just exactly who your Fula2 are but I missed the reply. EDIT Fula are spread from Senegal to Saudi. Your Fula2 are genetically Soninke.I just wanna know the precise ethny or locale of the Fula2 samples.
quote:Originally posted by Tukuler: Here are Laz2014 TreeMix 0:5
3 looks untenable but check the node where Mbuti splits off down to where everything else splits.
The 1st migration arrow is to Stuttgart. But where is the nock? Anyone?
In Figure S16.1 and 16.2 the source of the BE arrows in all cases seems to be somewhere close to where ancestral Mbuti originates. In Figure S16.3 and 16.4 the source of the BE arrows is literally basal to Eurasians. What is the answer you're looking for? I can say it's basal to Eurasians, but that would be me being Captain Obvious. Seems like you're after something deeper.
I find it interesting how the inclusion of chimp (if that's what did it) seems to change where treemix sees BE originating. It's like the chimp sample introduces important phylogenetic information that clarifies relationships.
@Tyrannohotep. Might be a good idea to go over these treemix runs again for your project.
quote:Originally posted by Tukuler: I tried to find out just exactly who your Fula2 are but I missed the reply. EDIT Fula are spread from Senegal to Saudi. Your Fula2 are genetically Soninke.I just wanna know the precise ethny or locale of the Fula2 samples.
They are all from Gambia in terms of location, Interms of ethny I can only provide what GGVP says they are, which is fulani.
Posts: 1782 | From: New York | Registered: Jul 2016
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
Thx EM Went there. Used the search. Couldn't find Serehule?
Thx Swenet I completely overlooked 16.1-2. Complicated now that I seen them. Bonzo and Cheeta have long been our best buds. I'm OK with them in analysis.
I think Laz's TreeMix wasn't examined before? Other than the, to me, improbability of the 3 migrations tree I really have no ideas. OK I think the nock is a E Med farmer. and I think the long slide between Mbuti and all other humans represents everything from Africans Africans basal to Eurasians / Eurasians basal to Africans 1st Mid Easterners / 1st successful Outside of Africa folk [not relying on the Remaud Laz Kelso Reich text]
I dunno.
So ask for ideas. Can't think of everything myself.
EDIT According to Laz 16.1-2 are "affected by ascertainment bias due to how the SNPs were chosen for the array."
16.3-4 whole genome sequencing overcomes that limitation and has greater statistical power.
quote:Originally posted by Swenet: In Figure S16.1 and 16.2 the source of the BE arrows in all cases seems to be somewhere close to where ancestral Mbuti originates. In Figure S16.3 and 16.4 the source of the BE arrows is literally basal to Eurasians. What is the answer you're looking for? I can say it's basal to Eurasians, but that would be me being Captain Obvious. Seems like you're after something deeper.
I find it interesting how the inclusion of chimp (if that's what did it) seems to change where treemix sees BE originating. It's like the chimp sample introduces important phylogenetic information that clarifies relationships.
@Tyrannohotep. Might be a good idea to go over these treemix runs again for your project.
I've looked at the TreeMix diagrams and have indeed noted the trend you described. The effect that including chimpanzee data had on the origin point of the BE->Stuttgart arrow is indeed quite curious. This result must be more accurate with regards to BE's actual affinity to OOA versus Mbuti, but it's funny that BE's African origin seems much more apparent when you exclude the chimp genome.
posted
@Tukuler Everything else seems to be mostly unaffected in the human origins dataset, so I thought rooting the tree with chimp did it. This is one of those times where having capable in-house TREEMIX hobbyists to redo the analysis would be helpful.
@Tyrannohotep Notice also the TREEMIX residuals. There is always some extra Mbuti affinity to the farmers that is not shown in the TREEMIX trees. WHG also has extra affinities to Africans as I already suggested earlier in this thread (see TREEMIX residuals for that, too).
Posts: 8785 | From: Discovery Channel's Mythbusters | Registered: Dec 2009
| IP: Logged |
posted
As I suspected, their only source is Lazaridis
Posts: 42940 | From: , | Registered: Jan 2010
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
For some reason I thought unsupervised meant no user chosen root bit I guess not. So Mbuti is the HOD root, Chimp the WGSD root.
And here I been thinking that acute angle was the root; Homo ancestor in HOD, Hominini ancestor in WGSD.
I think with both a spirit of tolerance and a strict no unproven nonsense and an enforced no sympathy non-trolling policy when it comes to science related threads this particular ES forum might attract new knowledgeable layman with tools.
Residuals are as important as the trees and arrows because they "identify aspects of ancestry not captured by the tree". But knowing what the colors show is not intuitive. The black and blue do mean a definite admixture, right? White is no admixture at all and red is ??? Nixing Cheeta that is.
Residuals on HOD show sharing between Mbuti with Iceman Stuttgart, Onge, Loschbour LaBrana.
WGSD with Bonzo and swapping Dai for Onge: Dai, Loschbour LaBrana, Karitiana, Stuttgart.
Shall we overinterpret?
quote:Originally posted by Swenet: @Tukuler Everything else seems to be mostly unaffected in the human origins dataset, so I thought rooting the tree with chimp did it.
@Tyrannohotep Notice also the TREEMIX residuals There is always some extra Mbuti affinity to the farmers that is not shown in the TREEMIX trees. WHG also has extra affinities to Africans as I already suggested earlier in this thread (see TREEMIX residuals for that, too).
posted
SUPERVISED simply means that the researchers chose(not the computer) the specific SNPs to do the comparison between populations. In other words it is NOT random. They may be as little as 100 SNPs or as much as several million.
That is why in some studies you will see MButi carrying "Eurasian" ancestry while in other studies they do NOT. It depends on what SNPs are chosen for the analysis. ie Supervised or unsupervised.
That is why you need to read the methodology section to understand WHAT they are doing. You cannot just read the results and infer.
eg in the Abusir papers if I understand the methodology section correctly they used a few SNPs on one chromosome, #19, which is highly unusual. Novice reader did not get that! That is why the Abusir paper is "fake". They tried to isolate few SNPs found is Yorubans to make their point. Then used YRI to represent ALL SSA.
Undersand the game ...and methodology.
-------------------- 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
Thx Mensa for the heads up for those who maybe didn't read SI 16 "An advantage of tree mix and mix mapper is that they are unsupervised procedures and hencet are less vulnerable to the concern that the prior expectations about human history of the researchers using them will bias the results."
In this supplement 265,521 sites were TreeMixed.
quote:Originally posted by Tukuler: Thx EM Went there. Used the search. Couldn't find Serehule?
Doesn't really matter. Came to see that Fula2 also strongly resemble Serere and Wolof, whom at first I didn't hastily see in your graph.
A few aDNA pops wouldn't fit neatly in the main redux colulmn. I'm satisfied with this for now.
One way to crack this nut: 1 Pick a color. 2 Note it's majority or plurality 'population'. 3 Follow the color through the populations. 4 prepare for surprises
I don't know why postimg.cc blurred the image making it useless in zoom, Photos, or Paint. Dammit! Maybe next build.
Any possibility of K=20 in small hi def chunks I can work with and reassemble in redux order? I mean me and you can fox with what's there but it's too much work for anybody else in breaking down the overview for pop names and colors narrower than 5%.
Also any suggested 'materials & methods for a caption text?
posted
The picture was probably too big. You'll have to probably post smaller parts.
quote:They tried to isolate few SNPs found is Yorubans to make their point. Then used YRI to represent ALL SSA.
YRI weren't the only ones used. You've been asked before: What SSA population did they neglect to use that they should have?
Posts: 2508 | From: . | Registered: Nov 2011
| IP: Logged |
posted
?? You do realize this is above your pay grade? It is not about what populations they used but what SNPs they chose to use.
This discussion don't come with crayola
-------------------- 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
We are back to the Hadza/Sandawe Tanzania Malawi origin of the Ancient Egyptian.
But based upon the SNP chosen(supervised), the Bedouins are the best Representation of the Abusirs(Africans in the near East). So when they say "Near East" they are talking geography not people. Understand the gamesmanship.
They are not talking Syrians and Palestinians etc.
We know the Bedouins are remnants of an African population in the Levant
-------------------- 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 |
quote:Originally posted by Tukuler: For some reason I thought unsupervised meant no user chosen root bit I guess not. So Mbuti is the HOD root, Chimp the WGSD root.
And here I been thinking that acute angle was the root; Homo ancestor in HOD, Hominini ancestor in WGSD.
I think with both a spirit of tolerance and a strict no unproven nonsense and an enforced no sympathy non-trolling policy when it comes to science related threads this particular ES forum might attract new knowledgeable layman with tools.
Residuals are as important as the trees and arrows because they "identify aspects of ancestry not captured by the tree". But knowing what the colors show is not intuitive. The black and blue do mean a definite admixture, right? White is no admixture at all and red is ??? Nixing Cheeta that is.
Residuals on HOD show sharing between Mbuti with Iceman Stuttgart, Onge, Loschbour LaBrana.
WGSD with Bonzo and swapping Dai for Onge: Dai, Loschbour LaBrana, Karitiana, Stuttgart.
Shall we overinterpret?
quote:Originally posted by Swenet: @Tukuler Everything else seems to be mostly unaffected in the human origins dataset, so I thought rooting the tree with chimp did it.
@Tyrannohotep Notice also the TREEMIX residuals There is always some extra Mbuti affinity to the farmers that is not shown in the TREEMIX trees. WHG also has extra affinities to Africans as I already suggested earlier in this thread (see TREEMIX residuals for that, too).
Or maybe some of the existing members can try their hand at it. Some members may wonder how they can contribute to bio-anthro conversations but don't know how. Some are just not well-read enough to feel comfortable making more contributions in genetics conversations. Running tests is good way to compensate. It's fairly routine. Once you understand it, you can keep doing it. I would do it myself, but I'm already more committed to bio-anthro than I want to. Plus, I've been able to get by all this time without learning the tools so I'm not interested enough to make the investment.
I hope new generation(s) of ES members will mature to at least some bio-anthro specializations and niches. Attracting people who already know how to do it is great but if you attract people from the blogs, they won't necessarily be sympathetic to African OOA migrations and things we already take for granted.
Good observations on the TREEMIX residuals. In a lot of cases TREEMIX is more reliable and better than f3 and f4. Excess derived allele sharing in f3 or f4 is not very informative. TREEMIX tells you where it comes from and it has ways of letting you know when all the samples have African admixture, as opposed to just the sample under consideration. For instance, excess allele sharing of Africans with population A does not mean population B is not also admixed with Africans. Population B may even have more African ancestry than population A, even though f4 or f3 may point to population A being closer to Africans. The Mbuti affinities with WHG in TREEMIX shows how TREEMIX can do more than just mechanically poop out the answer that farmers are closer to Africans than to WHG. TREEMIX shows scenarios are thinkable in which WHG may be closer to some Africans than farmers are.
Posts: 8785 | From: Discovery Channel's Mythbusters | Registered: Dec 2009
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
And just think if someone with the resource to download the tools and access databases inserted Mende and or Mbo, even Biaka is a better choice if a Shorty must be used.
Reduced to Mbuti and weighted by infusion sorry, no graphic designer skills or tools
Residuals on HOD show sharing between Mbuti with Iceman Stuttgart, Onge, Loschbour LaBrana.
WGSD with Bonzo and swapping Dai for Onge: Dai, Loschbour LaBrana, Karitiana, Stuttgart.
quote:Originally posted by Swenet: Everything else seems to be mostly unaffected in the human origins dataset, so I thought rooting the tree with chimp did it.
existing members can try their hand at it. Some members may wonder how they can contribute to bio-anthro conversations but don't know how. Some are just not well-read enough to feel comfortable making more contributions in genetics conversations.
Running tests is good way to compensate. It's fairly routine. Once you understand it, you can keep doing it. I would do it myself, but I'm already more committed to bio-anthro than I want to. Plus, I've been able to get by all this time without learning the tools so
I'm not interested enough to make the investment.
I hope new generation(s) of ES members will mature to at least some bio-anthro specializations and niches. Attracting people who already know how to do it is great but if you attract people from the blogs, they won't necessarily be sympathetic to African OOA migrations and
things we already take for granted. [not sure anymore]
Good observations on the TREEMIX residuals. In a lot of cases TREEMIX is more reliable and better than f3 and f4. Excess derived allele sharing in f3 or f4 is not very informative. TREEMIX tells you where it comes from and it has ways of letting you know when all the samples have African admixture, as opposed to just the sample under consideration. For instance,
excess allele sharing of Africans with population A does not mean population B is not also admixed with Africans. Population B may even have more African ancestry than population A, even though f4 or f3 may point to population A being closer to Africans.
Only La Brana and Loschbour are in the black. WHG also holds 4 out of 7 reaching blue-black. From all 12 residuals EEF only leads in 2 and isn't in 5 of them at all.
Residuals show ties the tree can't show. They need 'reconciliation' if synthesis isn't enough. Are residuals migrations without arrows? What does a non-phylogeny connection between 'Mbuti' and the 'Andaman Shorties' and nearby SE Asians mean? Let's just forget the America's Karitiana Mbuti link.
Man, I gonna go double take some old TreeMix pubs.
quote:Originally posted by xyyman: ?? You do realize this is above your pay grade? It is not about what populations they used but what SNPs they chose to use.
This discussion don't come with crayola
LMAO you are talking about "crayola" while showboating on that sh!tty MSPaint job? Thank you for the laugh, I needed that today.
In the OP I make clear my inexperience and I don't think those of us who're less experienced should feel bad about posting these threads or expressing a need for you to expand on a point in the hope it'll make sense to us at some point. The picture you post is still somewhat confusing. Ethiopian Jews have the brown component, yes. But the point of the study was that when looking at all the components together they resemble Arabs. Look, everybody's got things they're still learning to do. If you can't learn to photoshop when little kids not a fifth your age are doing it all over social media, please don't act like you're in any position to look down on other people at ES over perceptions of experience.
Posts: 2508 | From: . | Registered: Nov 2011
| IP: Logged |
posted
Oshun, xyyman's crayola inspired red lines are better than what he used to do, which was to use crisp ovals and graphic lines, sometimes people can get confused by these things and not know if they are part of the original chart or not
He knows how to make neat graphics > but it is better not to have those skills used on these charts so other people viewing these things know
Posts: 42940 | From: , | Registered: Jan 2010
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
Talk about off-topic and provocation?
It's good to be the queen!
quote:Originally posted by the lioness,: Oshun, xyyman's crayola inspired red lines are better than what he used to do, which was to use crisp ovals and graphic lines, sometimes people can get confused by these things and not know if they are part of the original chart or not
He knows how to make neat graphics > but it is better not to have those skills used on these charts so other people viewing these things know
posted
Man, you must have a bitch in the school yard!?
No CXXT. They do NOT resemble Syrians and Palestinians etc. They resemble Bedouins And Yemenis….who lives in the Near East. Don’t you get that?? Do you know the genetic history of the Bedouins and Yemenis……who live in Arabia? It is called geographic gamesmanship. It is trickery by the authors using the label Near East. Henn and others do the same thing. They choose an “Africanize” population in the Near East for their spin.(sometimes they blend in the Bedoiuns). Qataris (Henn) is about 60% African. 30% SSA and 30% NAian. There is a reason why Lazaridis chose The Bedouins as a proxy for EEF/BE which he could not disentangle from SSA.. Because they are North Africans, the first Neolithic Africans to migrate to the Levant. That is why the pre/Neolithic peoples in the Levant carry ancestral component for pigmentation. In other words they were black skinned. Stop arguing stuff you know nothing about. Stick to croyola?
And again you are missing the elephant in the room. It is not the Ethiopian Jews who are the kicker but the Sandawe!!!!! Foool!!! Even the South African Khoi-San carry the “brown” component. So deep in Sub-Saharan African genes related to the Abusir are found!!!! And in case you missed it the “red” component is found in late Neolithic Europe. Which mean the “red”/ Yoruban component was in Europe thousands of years before it hit Abusir. What does it all mean, an intelligent person will ask? How can the red component be outside African in Europe so long before entering Egypt(Abusir)? Your answer is …..?
BTW – Capra, ElMaestro and others can help me out here. What is the red component? I had to track down to 2 studies(Methodology) and I came back with a few components on Chromosome #19. Correct me?!
------------------------------ LMAO you are talking about "crayola" while showboating on that sh!tty MSPaint job? Thank you for the laugh, I needed that today. [Smile]
In the OP I make clear my inexperience and I don't think those of us who're less experienced should feel bad about posting these threads or expressing a need for you to expand on a point in the hope it'll make sense to us at some point. The picture you post is still somewhat confusing. Ethiopian Jews have the brown component, yes. But the point of the study was that when looking at all the components together they resemble Arabs. Look, everybody's got things they're still learning to do. If you can't learn to photoshop when little kids not a fifth your age are doing it all over social media, please don't act like you're in any position to look down on other people at ES over perceptions of experience. ------------------------------------
Oh! And Lioness is correct. I work in R&D type field. But I like my mock-up charts now. It is more effective to highlight my point.
-------------------- 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
@Sage. Nice to see you are doing a deep dive into TreeMIx. More power to you. I wish I had the time to play around with these software that can really tie-in population relatedness. ADMIXURE cannot. Treemix and GPS etc can. But most are not Windows based.
------------ Only La Brana and Loschbour are in the black. WHG also holds 4 out of 7 reaching blue-black. From all 12 residuals EEF only leads in 2 and isn't in 5 of them at all.
Residuals show ties the tree can't show. They need 'reconciliation' if synthesis isn't enough. Are residuals migrations without arrows? What does a non-phylogeny connection between 'Mbuti' and the 'Andaman Shorties' and nearby SE Asians mean? Let's just forget the America's Karitiana Mbuti link.
Man, I gonna go double take some old TreeMix pubs. ________________
I remember I had that discussion with Davidski on his blog. I challenged him to use TReeMix on the Abusir mummies with certain African groups. He was being his usual arrogant self, stating it would not make a difference but he would run it. A few days later he came back a deleted all my comments about the subject. Guess he did not like the results.
-------------------- 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 |
quote:Originally posted by xyyman: SUPERVISED simply means that the researchers chose(not the computer) the specific SNPs to do the comparison between populations. In other words it is NOT random. They may be as little as 100 SNPs or as much as several million.
That is why in some studies you will see MButi carrying "Eurasian" ancestry while in other studies they do NOT. It depends on what SNPs are chosen for the analysis. ie Supervised or unsupervised.
That is why you need to read the methodology section to understand WHAT they are doing. You cannot just read the results and infer.
eg in the Abusir papers if I understand the methodology section correctly they used a few SNPs on one chromosome, #19, which is highly unusual. Novice reader did not get that! That is why the Abusir paper is "fake". They tried to isolate few SNPs found is Yorubans to make their point. Then used YRI to represent ALL SSA.
Undersand the game ...and methodology.
I find it unnerving that this (unsupervised vs Supervised ) has been explained to you a number of times and you still choose to not pay attention and further mislead the forum.
The researchers ALWAYS choose the SNP Panel. Its never random. If it was "random" it would be very likely that ALL humans would have the same SNP's when you choose a random 500 thousand out of a few BILLION base pairs. This SNP Array is usually the Tens of Thousands, hundreds of thousands or even Millions of SNP's used in the ADMIXTURE analysis. They tell you exactly how many SNP's are used in the analysis. Different genetic companies have mostly pre-selected these SNP's for a specific reason: These are the ones that differ the most in sampled human populations. Companies like Illumina and Affymetrix come to mind.
SUPERVISED - Is when the HUMAN researcher pre-selects the human population samples to act as representatives of a specific K cluster.
UNSUPERVISED - Is when the program create homogeneous components on ITS OWN based on the specific K=Number.
IN an example, lets say you have Mandinka, Yoruba, French, Japanese and South Indian. K=4 UNSUPERVISED - may give you West African, European, South Indian and East Asian Clusters. My ancestry would come up as 88% West African, 11% European and 1% East Asian
But what if you were looking to analyze the African specific ancestry of African American samples? The unsupervised run would not be sufficient. You would then choose to SUPERVISE the admixture analysis and FORCE the splitting of Mandinka and Yoruba into 2 separate components of the K=4. You would then tell the computer French and Japanese are the other two.....that would leave South Asian as the "odd Man out".
My Ancestry would then be 50% Yoruba, 38% Mandinka 11% French and 1% Japanese.........perhaps giving a better narrative of what you are trying to research if lets say your study was on the African source origins of African descendant new world populations.........and since African Americans were the focus perhaps you wouldn't care that Southern Indians were left as a composite of French and Japanese (Just as nobody cares that Somali are modeled as 55% Yoruba and 45% Natufian).
The difference between unsupervised vs Supervised is the discretion of the Researcher vs the discretion of the Computer running the program. GENERALLY.....the researcher will run an unsupervised test first....see what comes out...nitpick/cull/prune certain individuals in a sample and then run a Supervised test based on some relatively "pure" (Homogeneous) representatives from the unsupervised test.
Posts: 2463 | From: New Jersey USA | Registered: Dec 2007
| IP: Logged |
posted
@Tukuler You can use the pdf for a higher resolution, however the image is huge. If you want me to replot the graph you'll have to wait as that gets added to the to do list.. I still haven't gotten around to processing the Taforalt genomes... as you can see, I've been posting raw data as of late, due to time constraints.
In regards to materials, I scraped together populations from various projects, Laz, AGVP, skoglund... For methods, I removed relatives from each dataset, thinned/prunned the dataset to avoid LD bias, and ran this dataset through ADMIXTURE. This dataset is meant to highlight African variation specifically in the presence of unique ancient genomes. So for the most part looking at this will let us know with a higher level of confidence which African populations truly have Eurasian Admixture and to what extent. The draw back is Eurasian diversity as E.Eurasians-HGs and West Eurasians remain joined together due to how relatively homogenous OOAs are. This is unsupervised btw.
Best statistical run is at K7, there's an unusual dip in CV for K10 when pemba and Mota get their own components, which makes sense. K9 & K13 is where the elusive North east african component arrives.
Posts: 1782 | From: New York | Registered: Jul 2016
| IP: Logged |
posted
I will get back to you Beyoku. You and your misguided rants.
anyways...
Who were the Nataruk people? Mandibular morphology among late Pleistocene and early Holocene fisher-forager populations of West Turkana (Kenya) Aurélien Mouniera, b, , , Maria Correiab,
Abstract Africa is the birthplace of the species Homo sapiens, and Africans today are genetically more diverse than other populations of the world. However, the processes that underpinned the evolution of African populations remain largely obscure. Only a handful of late Pleistocene African fossils (∼50-12 Ka) are known, while the more numerous sites with human fossils of early Holocene age are patchily distributed. In particular, late Pleistocene and early Holocene human diversity in Eastern Africa remains little studied, precluding any analysis of the potential factors that shaped human diversity in the region, and more broadly throughout the continent. These periods include the Last Glacial Maximum (LGM), a moment of extreme aridity in Africa that caused the fragmentation of population ranges and localised extinctions, as well as the ‘African Humid Period’, a moment of abrupt climate change and enhanced connectivity throughout Africa. East Africa, with its range of environments, may have acted as a refugium during the LGM, and may have played a critical biogeographic role during the heterogene`ous environmental recovery that followed. This environmental context raises a number of questions about the relationships among early Holocene African populations, and about the role played by East Africa in shaping late hunter-gatherer biological diversity. Here, we describe eight mandibles from Nataruk, an early Holocene site** (∼10 Ka)** in West Turkana, offering the opportunity of exploring population diversity in Africa at the height of the ‘African Humid Period’. We use 3D geometric morphometric techniques to analyze the phenotypic variation of a large mandibular sample. Our results show that (i) the Nataruk mandibles are most similar to other African hunter-fisher-gatherer populations, especially to the fossils from Lothagam, another West Turkana locality, and to other early Holocene fossils from the Central Rift Valley (Kenya); and (ii) a phylogenetic connection may have existed between these Eastern African populations and some Nile Valley **and Maghrebian **groups, who lived at a time when a Green Sahara may have allowed substantial contact, and potential **gene flow**, across a vast expanse of Northern and Eastern Africa.
-------------------- 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 |
quote:Originally posted by Elmaestro: [Q] @Tukuler You can use the pdf for a higher resolution, however the image is huge. If you want me to replot the graph you'll have to wait as that gets added to the to do list.. I still haven't gotten around to processing the Taforalt genomes... as you can see, I've been posting raw data as of late, due to time constraints.
In regards to materials, I scraped together populations from various projects, Laz, AGVP, skoglund... For methods, I removed relatives from each dataset, thinned/prunned the dataset to avoid LD bias, and ran this dataset through ADMIXTURE. This dataset is meant to highlight African variation specifically in the presence of unique ancient genomes. So for the most part looking at this will let us know with a higher level of confidence which African populations truly have Eurasian Admixture and to what extent. The draw back is Eurasian diversity as E.Eurasians-HGs and West Eurasians remain joined together due to how relatively homogenous OOAs are.***This is unsupervised btw.***
Best statistical run is at K7, there's an unusual dip in CV for K10 when pemba and Mota get their own components, which makes sense. K9 & K13 is where the elusive North east african component arrives. [/Q]
Why would you remove LD in the methodology? That is the whole point isn’t it. To show Affinity. That is only done through LD or STRs..isn’t it?
Quote: “... For methods, I removed relatives from each dataset, thinned/prunned the dataset to avoid LD bias, and ran this dataset through ADMIXTURE. This dataset is meant to highlight African variation specifically in the presence of unique ancient genomes.”
-------------------- 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
Now back at Beyoku. No disrespect but sometimes you come across dumber than bricks…like your buddy Oshun. Why is it always a pissing contest with you. You just agreed with me and in your mind you are making a point to debunk mine. I said you are NOT in the same league as me….Why is that so difficult to understand Ms Wannabe?
Again..by YOUR definition
SUPERVISED - Is when the HUMAN researcher pre-selects the human population samples to act as representatives of a specific K cluster.
UNSUPERVISED - Is when the program create homogeneous components on ITS OWN based on the specific K=Number.
Unsupervised=computer program makes or choses the SNPs Supervised = is when humans chose the SNPs.
Supervised=manipulation by the researchers as with the Abusir chosing a few SNPs on Chromosome #19...if I traced down the methodology correctly. As I said I needed to go through 2 additional researcher papers to understand what they did with the Abusir. The manipulation they conned people with?
We saw another good example of manipulation with the “no SSA in Natufians” which later changed to 6-29% when the lie was exposed.
To the newbies. How is the manipulation done? The researcher can select ONLY 50 out of 50 million SNPs and tell the computer to find the population that has these 50!! SNPs. And of course they will chose 50 SNPs NOT found in Nigerians because they have the entire genome of YRI. But left up to the computer (supervised) the computer will find 25 million SNPs found in BOTH Nigerians and Abusir but not in Europeans. Understand the game of supervised vs unsupervised!! Even unsupervised can be manipulated but that is todays lesson. (wink) Understand the methodology so you cannot be fooled.
That is why in unsupervised cluster charts (Rosenberg et al 2002) Northern Europeans are 60% African vs non-African and Southern Europeans are 80% African vs non-African.
Lazaridis confirmed that 10years later by proclaiming southern Europeans are 80% EEF/African.
Don't get twisted with the terminology.
Are Europeans 80% Tanzanian Hadza/Sandawe? This can be a bitch!!
FLOODING: combined like posts //MOD Stop giving me excess work Xyyman
-------------------- 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
@xyyman... in very simple terms closer genes are together, the higher % chance they'll travel together. ADMIXTURE determines the probability certain genes came together and from whom simultaneously. So in theory you'd want variants to be more neutral to avoid genes that are in high LD from auto assigning to a particular ancestor using this tool.
Weighted LD decay tests are an entirely different way to help determine admixture. you assign a potential recipient and two donors and compare the difference of the frequency of each pair of variants in the donors. And multiply that with the rate at which the pair of variants occur together in the recipient population. Each pair of genes will yield a weighted score that correlates with the distance they are from each other.
Posts: 1782 | From: New York | Registered: Jul 2016
| IP: Logged |
quote:Originally posted by the lioness,: Oshun, xyyman's crayola inspired red lines are better than what he used to do, which was to use crisp ovals and graphic lines, sometimes people can get confused by these things and not know if they are part of the original chart or not
Something like this:
Is not hard to understand. Which basically means even if he does know how to technically use photoshop to some degree, he still doesn't understand the principles of graphic design very well to use it. Shit the markups on the Irish study in the OP looked better. I grasp other people here may also use MSPaint but they don't make every other post about how experienced they are and how everyone wants to be them while they spend endldess threads talking to themselves. He is in no position to complain about experience. Everyone has things they need to learn. And if he doesn't like this thread he can go back to humming and mumbling to himself again in his own thread.
quote:Originally posted by xyyman: Now back at Beyoku. No disrespect but sometimes you come across dumber than bricks…like your buddy Oshun.
Please nevermind him Beyoku, he's attention starved because no one is giving him enough to do on the Phoenecian thread he started. If you take a peek in there, he's pretty much talking to himself. But if you must respond, please just discuss the parts about supervised/unsupervised for those of us who want to have more of an understanding on how it works.
Posts: 2508 | From: . | Registered: Nov 2011
| IP: Logged |
Are Europeans 80% Tanzanian Hadza/Sandawe? This can be a bitch!!
yes, Southern Europeans are as African as African Americans but they're afraid to admit it
Posts: 42940 | From: , | Registered: Jan 2010
| IP: Logged |
----- @xyyman... in very simple terms closer genes are together, the higher % chance they'll travel together.
EXACTLY!!!! That is why linked loci ie STRs can be used to determine geographic affinity. That is why the Amarnas came back Sub-Saharan African and the Abusir came back Near East. They did not used LD type alleles for the Abusir
ADMIXTURE determines the probability certain genes came together and from whom simultaneously.
No! If I understand it correctly ADMIXTURE determines **shared** ancestry. While TreeMix uses the same SNPs from ADMIXTURE to determine DIRECTION of migration and the likely events
So in theory you'd want variants to be more neutral to** avoid** genes that are in high LD from auto **assigning to a particular ancestor** using this tool. ditto above
Weighted LD decay tests are an entirely different way to help determine admixture. you assign a potential recipient and two donors and compare the difference of the frequency of each pair of variants in the donors. And multiply that with the rate at which the pair of variants occur together in the recipient population. Each pair of genes will yield a weighted score that correlates with the distance they are from each other.
?? are you BSing me ElMaestro? razzle dazzle? WTF is weighed LD tests?
-------------------- 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 |
Xyyman... read This for info on ADMIXTURE (Pay attention to page 4)
And this for using LD to infer admixture events and dates
...I'm not going to try to explain this anymore.
@Oshun I believe Lioness' point is that it's hard to distinguish whether or not the mark up is authentic when done so "professionally." To her point, the example you give is actually marked up in the source iirc. So in order to not look like he's posting "doctored" charts etc. he's better off working sloppily.
Posts: 1782 | From: New York | Registered: Jul 2016
| IP: Logged |
quote:Originally posted by Elmaestro: @Oshun I believe Lioness' point is that it's hard to distinguish whether or not the mark up is authentic when done so "professionally." To her point, the example you give is actually marked up in the source iirc. So in order to not look like he's posting "doctored" charts etc. he's better off working sloppily.
I get her point. My point however is that while it can be difficult at times, it's not impossible to make a neat markups without mistaking it. Does it require skill in graphic design beyond technical aptitude to use the program? Often times yes. But is it impossible? No. Yes the paper is marked in the material released online, but as part of the review process. Obviously the original document hadn't intended to make the errors marked up.
In any case, those markups aren't confusing. "Mastery" of graphic design involves understanding how to avoid those pitfalls and that he has more to learn on the issue. Not a problem, but he seems to like talking sh!t when people acknowledge they have plenty to learn and are willing to question, but doesn't like looking in his own backyard.
quote:Originally posted by xyyman:
?? are you BSing me ElMaestro? razzle dazzle? WTF is weighed LD tests?
It's not necessary to be that defensive. Relax it's not that serious, everyone has stuff they don't know. Posts: 2508 | From: . | Registered: Nov 2011
| IP: Logged |
Tukuler
multidisciplinary Black Scholar
Member # 19944
posted
quote:Originally posted by Elmaestro: You can use the pdf for a higher resolution, however the image is huge.
You know good and damn well you shoulda linked that the 1st time stead of some damn 304kb png. Only resourceless me can do png.
quote: If you want me to replot the graph
Not sure what you mean here.
quote: In regards to materials, I scraped together populations from various projects, Laz, AGVP, skoglund...
For methods, I • removed relatives from each dataset, • thinned/prunned the dataset to avoid LD bias, and • ran this dataset through ADMIXTURE.
So for the most part looking at this will let us know with a higher level of confidence which African populations truly have Eurasian Admixture and to what extent.
My focus is Africa. Proving to what extent Euras got it off with Africans? Not me. African to African and African to non-African is where I'm coming from.
In my book Africa ends where Turkey's and Iran's mountains begin. No nevermind Neolithic Anatoli @ 48% in Natufian. It's their 52% African with Bataheen, S Afr2000BP, Mota, and Amazigh, (like).
Geography, climates, biomes, bones, bric-a-brac, myths, legends, primary documents, and history inform me on Afer <-> Eura interminglings. Don't mean we can't work together. Better works come from wider input, yes?
quote: This is unsupervised btw. Best statistical run is at K7,
Cool, but thanks you went to K=13. Couldn't've read Natufian w/o it. It also implies there's a African in the HOtTUb with the CHi-town Gangster.
quote: there's an unusual dip in CV for K10 when pemba and Mota get their own components, which makes sense. K9 & K13 is where the elusive North east african component arrives.
Pemba700 I presume?
Incredible multi component Malawi 6000BP with 3500 year apparent continuity. Where's it in other studies since first published? Seems a prime, must use population just like you see. Is it because of it's pluralities?
Maximized those pluralities are • S Afr 2000BP • Ju/'Hoansi • Nuba • its own • Mota • Herero at the K=13 level. Six way substructure in ~4000 BCE Fingara Malawi. Southeast Africa foundational homeland of southern to NE Afr peoples? Random Gathering of the Tribes from Parts Unknown? Some combination of the two? Something else? ???
When I redux your PDF I'm a open a thread for comment on your work. Don't our own members deserve what we give the pros?
Meanwhile I'd still like to read what all y'all out there thinking.
posted
Several years ago the Lioness was gaffing at my comment on Henn proclamation that the LWK (Kenyans)were ancestral to Maghrebians/Europeans . Now we have narrowed it down further. Not the LWK but Tanzanians Hadza and Sandawe are really the Ancestral group to “Eurasians”. Who would have thought? Sandawe and Hadza carry more European ancestry than African-Americans
-------------------- 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 |
[QUOTE]Originally posted by Elmaestro: [Q] Well, you don't understand **correctly**,
-------------------- 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 |
Are Europeans 80% Tanzanian Hadza/Sandawe? This can be a bitch!!
quote:Originally posted by xyyman: Several years ago the Lioness was gaffing at my comment on Henn proclamation that the LWK (Kenyans)were ancestral to Maghrebians/Europeans . Now we have narrowed it down further. Not the LWK but Tanzanians Hadza and Sandawe are really the Ancestral group to “Eurasians”. Who would have thought? Sandawe and Hadza carry more European ancestry than African-Americans
Some have suggested Tanzanians are ancestral to the Nubians and ancient Egyptians. So does this mean Southern Europeans are closer to AE than West Africans?
Posts: 42940 | From: , | Registered: Jan 2010
| IP: Logged |
posted
No Lioness! You need to understand which came first “the chicken or the egg”.
First off the FACT is all humans are related. @K2 in an unsupervised cluster chart Europeans are more African than non-African. Even @K2 ***ALL*** Africans carry non-African ancestry. The question is what and when are the migration events? That is where TreeMix and the like comes in. Capra likes to point the separation was about >60K ya but we know what he is about and his motive But Skoglund and Lazaradis studies has thrown everything out of sort. Because 6000years ago Europeans(WHG) did NOT carry Basal Eurasian. Europeans back then were Dravidians/Onge/Melanesians(Mathieson et al)…That is a FACT. 8000years ago BE was in deep Sub-Saharan Africa in Malawi Hora and Luxmanda. This is what the data shows.
So obviously Europeans are primarily Neolithic as Lazaridis pointed out. But before BE hit Southern Europe from Tanzania via the Nile and Green Sahara it had to have existed in the Sahara maybe for thousands of years before entering S. Europe/Near East , ala Taforalt/Natufians. As the aDNA has shown. So that is why Europeans have absolute no close genetic affinity to AEians compared to West Africans and Great Lakes and Southern Africans. West Africans are part of the genetic stock that remained on the continent….primarily…since they may carry Iwo Eleru DNA or some archaic West African homonid. Time will tell. Europeans had already left the continent and AEians are indigenous Africans as the STR has shown. STR are in LD and takes 1000’s of years to “form”. That why they cannot take random SNP(unsupervised) and try to draw geographic affinity. Understand the game.
quote:Originally posted by the lioness,:
quote:Originally posted by xyyman:
Are Europeans 80% Tanzanian Hadza/Sandawe? This can be a bitch!!
quote:Originally posted by xyyman: Several years ago the Lioness was gaffing at my comment on Henn proclamation that the LWK (Kenyans)were ancestral to Maghrebians/Europeans . Now we have narrowed it down further. Not the LWK but Tanzanians Hadza and Sandawe are really the Ancestral group to “Eurasians”. Who would have thought? Sandawe and Hadza carry more European ancestry than African-Americans
Some have suggested Tanzanians are ancestral to the Nubians and ancient Egyptians. So does this mean Southern Europeans are closer to AE than West Africans?
-------------------- 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 |
before BE hit Southern Europe from Tanzania via the Nile and Green Sahara it had to have existed in the Sahara maybe for thousands of years before entering S. Europe/Near East , ala Taforalt/Natufians. As the aDNA has shown. So that is why Europeans have absolute no close genetic affinity to AEians compared to West Africans and Great Lakes and Southern Africans.
Why would something existing in the Sahara for thousands of years not have any commonality with the Ancient Egyptians?
Posts: 42940 | From: , | Registered: Jan 2010
| IP: Logged |
“All humans are related” and have commonality. Native Americans “genes” were found in the Amarnas……….DNAConsultants.
The utility of short tandem repeat loci beyond human identification: Implications for development of new DNA typing systems(1999)
Notice THO1 repeats are very similar in Africans(AFRAMS) and Native American. That does not mean the Amarnas are Native American
Quote by DNAConsultants “Like most of the other genes in the family, it is Central African in ancient origin, but unlike the other markers it has a sparse distribution outside Africa with a worldwide average frequency of 4%. Still, Africans and African-influenced populations (1 in about 10) are about twice or three times as likely to have it as non-Africans”.
-------------------- 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
That is why we need the STR of the Abusirs. The Gov't don't use SNPs frequency or even ADMIXTURE. If they want to know if the criminal is an European they use pop STR. No other way around that!!
So Elmaestro and others. You get my point? You cannot “prune” out LD related genes. That is like removing STR which are essentially LD. They are transmitted as blocks of genes. That is why Berbers are Negros as I have said many times. Forget the visuals their genes tell the story. When someone has figured out how to pull the STR from the BAM files from the Abusir it will come back 100% AFRICAN just as the Amarnas.
In fact I will go out on a limb to proclaim the STR profile of La Brana and the STR profile of Toforalt North Africans will be very similar….why. It takes may be thousands of years for the STR profile of a population to be established. That is why DNATribes has the Amarnas being more Great Lakes Africans than West African. Like Europeans, West Africans had left the Nile Valley….albeit more recent than Europeans.
-------------------- 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
To those who don’t get it…..the Neolithics/EEF/Natufians originated somewhere close to Malawi/Tanzania in SSA about 10-15000 years ago. From there they traveled the super highway(down River of the Nile) and fanned out throughout the Sahara. They encountered more ancient humans. Iwo Eleru, La Brana related peoples etc. From the Sahara they entered southern Europe like Sardinia/EEF, Italy and Iberia….and the Levant and Near East. They also entered West Africa and maybe spread south into Southern Africa.
They split into Western Farmers(west Europe) and Eastern Farmers(Levant and Asia)….ie Natufians(East of the Nile) and EEF(west). That is why West Europeans are NOT related to Natufians irregardless to the bloggers. The researchers said so. Read the study! But BOTH Natufian and EEF genes are found in Tanzania. Not only the genes tell the story both the skulls. Sergi, Coon etc. Also read the thread by Sage. Black skin White Skull. This was common knowledge back in the 1800’s
-------------------- 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
How in the world are the Abusir going to be 100% African?! What, do you think they were as pure as the Dinka?
Posts: 1568 | From: Pluto | Registered: Sep 2008
| IP: Logged |
posted
^ Another dumb ass who don't have a clue. What you need to do is point out where they are NOT 100% pure African. Show me the data. No parroting now!!!
anyways...
Here is what the FBI has to say. ….
Quote: “5 SNP versus STR As mentioned in Section I, technological advances made in the context of the Human Genome Project mainly dictated the shift of DNA technologies that are being implemented in DNA forensics. The recent findings of the abundance of SNPs and the ease of automation and miniaturization of detection techniques (see [8, 36]) are already prompting introduction of microchip-based SNP assays for DNA forensic analyses
and asked how many such loci would yield the combined power of the random match probability offered by the 13 STR loci. Instead of using the exact values of the estimated random match probabilities for each of the seven world populations, we determined the number of SNP loci needed to reach random match probabilities of the range from 1 in a billion to 1 in 1000 trillion (the range of values observed in Table 3 for the 13 STR loci).
Hence, more careful validation studies of SNP loci would be needed before implementing them for forensic and paternity analysis. In addition, the efficiency of SNP loci for interpreting DNA mixture evidence is far **more reduced,** necessitating a far greater number of them to equal he potential of the present set of 13 STR loci.A detailed account of this will be discussed elsewhere.” ______________________________
Relax, you barely understand what you're talking about yourself. Removed (1) and will continue to delete flamebait posts //MOD
-------------------- 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
You're clearly the most brilliant man in here, and everyone pales in comparison. Demonstrate (via data) how Northeast Africans are now 100% African despite some populations being 40% Eurasian.
Posts: 1568 | From: Pluto | Registered: Sep 2008
| IP: Logged |
quote:Originally posted by sudaniya: You're clearly the most brilliant man in here, and everyone pales in comparison. Demonstrate (via data) how Northeast Africans are now 100% African despite some populations being 40% Eurasian.
xyyman says haplogroup H, U and J originate in Africa that's how
Posts: 42940 | From: , | Registered: Jan 2010
| IP: Logged |
posted
All genes have an African origin. Even the labeled "Eurasian"
-------------------- 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
@Tukuler Good job adding info to Lazaridis' TREEMIX analysis.
For those who may be wondering if WHG affinity with Africans is just because WHG absorbed farmer ancestry, this is not true. Or at least, it's not completely true. WHG have African ancestry that's much older than the arrival of farmers 7ky ago. We need to go back to postglacial times to explain some of the African ancestry in WHG.
quote:The emergence of the use of the becs of Tamar Hat is synchronous with the end of the Upper Paleolithic, prior to the expansion of bec use in the Upper Magdalenian of Europe, where technical and stylistic convergences were observed. This raises the question of the emergence of these lithic implements in North Africa and the possible spread of similar industries elsewhere.
quote:Originally posted by Swenet: @Tukuler Good job adding info to Lazaridis' TREEMIX analysis.
For those who may be wondering if WHG affinity with Africans is just because WHG absorbed farmer ancestry, this is not true. Or at least, it's not completely true. WHG have African ancestry that's much older than the arrival of farmers 7ky ago. We need to go back to postglacial times to explain some of the African ancestry in WHG.
quote:The emergence of the use of the becs of Tamar Hat is synchronous with the end of the Upper Paleolithic, prior to the expansion of bec use in the Upper Magdalenian of Europe, where technical and stylistic convergences were observed. This raises the question of the emergence of these lithic implements in North Africa and the possible spread of similar industries elsewhere.
UBBFriend: Email this page to someone!
![[Big Grin]](biggrin.gif)
![[Smile]](smile.gif)
![[Wink]](wink.gif)
Printer-friendly view of this topic