Talk:African admixture in Europe

Does the article NAME make sense?

I'd like to ask whether this article name makes sense. Admixture of what? Biodiversity? Languages? Music styles? If the subject is human genetic diversity then doesn't it need some more words?--Andrew Lancaster (talk) 11:26, 26 August 2009 (UTC)

I would prefer to hear from others before starting to move the article to a better name.--Andrew Lancaster (talk) 15:00, 28 August 2009 (UTC)

Suggestions,

• African DNA admixture in Europe
• African genetic admixture in Europe

Wapondaponda (talk) 16:37, 28 August 2009 (UTC)

• I agree the name is sub-optimal. African-European human genetic admixture is a bit of a mouthful, but is probably reasonable. One goal of a good article name is that it's easy for somebody to find it directly by knowing what to type into the search box. It's unlikely that's going to happen with this article no matter how the name gets tweaked; the only likely way I can see for somebody to get here is by clicking a link from a more generic article, dab page, category page, etc, so I'm not worried too much about the exact name being easy to guess/predict. I assume there are other articles on similar topics (Geographic area X admixture into geographic area Y), and it would be nice if there was a common naming scheme for all of them. -- RoySmith (talk) 16:53, 28 August 2009 (UTC)

Google gives some insights, and it reminds about some of the things we have to be careful of too. Admixture is often a word used to describe a sort of statistical simulation which is not always even meant to imply causality. It is a technical term I fear many people misuse. We are talking about good old fashioned "mixture" to speak, because if I understand the intention we ARE talking about real movements of people. For example when a study describes Ethiopians in terms of Middle Eastern and Sub Saharan "admixture" they are not necessarily meaning to say that Ethiopians descend from the mixture of two migrations which came from those places.

To use a bit of jargon there is a difference between Population Structure and Population History. Is this article not about the latter, i.e. "population movements"? Please opine.--Andrew Lancaster (talk) 17:03, 28 August 2009 (UTC)

Admixture originally refers to non genetic mixtures. However it seems to have recently entered the genetics lexicon. One more suggestion

• African gene flow into Europe.

Wapondaponda (talk) 06:53, 2 September 2009 (UTC)

Maybe better than what we have but...

• It treats genes as things that flow. I like to avoid such metaphors in this subject because people get confused enough already.
• No mention of humans.

African contributions to human genetic diversity in Europe is not catchy, but does it need to be catchy?--Andrew Lancaster (talk) 08:28, 3 September 2009 (UTC)

Hoping this question will not be forgotten.--Andrew Lancaster (talk) 10:01, 7 September 2009 (UTC)

Opening sentences

I'd like to go through the opening sentences and raise some issues I think need addressing...

African admixture in Europe refers to the European presence (via admixture) of human genetic lineages typical of the peoples of Africa including Sub-Saharan Africa.

First, this is a definition, presenting itself as a well-known one from outside Wikipedia. Is it one?

This evidence pertains to geneflow suspected after humans left Africa via Arabia more than 40,000 years ago.

Is it really appropriate to put this very specific migration down in the intro without any discussion of it? I think not. A casual reader will get the impression that there is no debate about this, and they'll also probably tend to guess why this migration is being distinguished. Will they think this was the first migration for example? I am not sure we need this sentence.

The evidence suggests that there has been gene flow between Africa and Europe in both prehistoric and historic times.[1]. In more recent history, the peoples of Europe and Africa came into contact during the, the exploration and colonisation of Africa and as a consequence of the Atlantic slave trade.[2] As a result of these recent contacts, lineages of African descent have also been detected in Europe. In general, African admixture is distributed along a South-to-North cline, with peaks in the Mediterranean region and Iberia.

How about...

This article summarizes what is known or proposed concerning the presence of African genetic lineages in European populations. The first anatomically modern humans in Europe are thought to have had African ancestry, and genetic evidence suggests subsequent movements from Africa to Europe in prehistoric and historical periods.

I think the rest is for the body of the article, and the article should be divided into time periods. I see no reason to exclude any time period from the first anatomically modern humans in Europe onwards.--Andrew Lancaster (talk) 11:39, 26 August 2009 (UTC)

I definitely agree that the lead needs some refinement. However I also find a brief timeline of events important as it helps to define admixture. The OOA migration is the time=0, the reference point. However, I am open to a suggestion that incorporates this information into the body of the article. Wapondaponda (talk) 22:57, 26 August 2009 (UTC)

And which OOA migration is "the" one? Is this concept really so simple that you can pass it off quickly in an intro?--Andrew Lancaster (talk) 05:55, 27 August 2009 (UTC)

Merger proposal

The issue of merging was specifically dealt with at Wikipedia:Articles for deletion/African admixture in Europe and the decision was to keep a separate article. So I will remove the tag. Wapondaponda (talk) 06:57, 2 September 2009 (UTC)

Please allow reasonable time for discussion with more editors. Abductive (reasoning) 08:24, 2 September 2009 (UTC)

Given how much discussion there has been already, including very recently, shouldn't you as the proposer give some explanation about what new issues you want to raise?--Andrew Lancaster (talk) 11:57, 2 September 2009 (UTC)

The subject of the AFD was specifically whether to have a separate article or merge into Genetic history of Europe. this comment specifically addresses the issue. I feel a discussion about merging is redundant, and means going through the process all over again. I agree with Andrew in that unless there is some new information, there is no need to go through what was discussed just last week. Wapondaponda (talk) 06:35, 3 September 2009 (UTC)

I assume that this is a content fork. Aren't you guys on opposite sides? Abductive (reasoning) 06:46, 3 September 2009 (UTC)

Andrew prefers not to take sides in disputes, so we are not on opposite sides or on the same side. Wapondaponda (talk) 06:59, 3 September 2009 (UTC)

Interesting. Still a content fork. Abductive (reasoning) 07:05, 3 September 2009 (UTC)

Not necessarily a CFORK, actually the genetic history of Europe article may have more problems than this article. Wapondaponda (talk) 07:08, 3 September 2009 (UTC)

But a better title. Abductive (reasoning) 07:14, 3 September 2009 (UTC)

The contesting content on Genetic History of Europe has been removed, therefore it is no longer a content fork. The better way to describe it is as divide and conquer, get this material of another page so that the other page can evolve, and this page, well we shall see if the editors take their role as wikipedians seriously or not.PB666 ^yap 16:21, 3 September 2009 (UTC)

My current feeling is that this article should be given a chance to see if editors can find enough non controversial and properly sourced material to make a non controversial and properly sourced article. The deletion debate about this article went "my way" in the sense that I argued there was enough of a change of subject to make it unclear whether this article would necessarily have the same problems as the last one. But actually I did not vote. I therefore believe it is critical that this article sets itself a clear title and direction. See the subject above, which I believe is not being handled with sufficient seriousness.--Andrew Lancaster (talk) 08:22, 3 September 2009 (UTC)

Agreed with AndrewPB666 ^yap 16:17, 3 September 2009 (UTC)

Work to do

Back to the main issue, there is actually a lot more information that will be added to this article. The autosomal markers, for which there is a lot more detail, are only briefly mentioned here. If we had to merge back, Genetic history of Europe will have too much information about Africa. Already there are suggestions to delete the admixture sections on Genetic history of Europe as there are already articles that deal with them, such as this article. Wapondaponda (talk) 07:59, 3 September 2009 (UTC)

Intro

I have changed the intro quite a lot in the way I already implied I thought necessary. Now the article starts by laying out the basics of what mainstream archaeology says about the chronological starting point. This gives a basis which can be built upon rather than leaving everything to vaguely implied assumptions. Leaves things between the lines might end up getting accused of being a way of quietly pushing a POV, and we now that such accusations do tend to fly around this subject!--Andrew Lancaster (talk) 09:25, 3 September 2009 (UTC)

COI heads up

Just a note to say I have cited myself in a trivial way. The article seemed to need a comment about the age of E1b1b1b (E-M81), but the primary literature is all over the place. Rather than taking a position myself (I think it is North African Neolithic more likely than not, but frankly there is very little to go on and it certainly probably split from the other major clades before the Holocene), we at least need a reference to a review saying that there are big variations in the approximations, and there is only one such review I think.--Andrew Lancaster (talk) 10:49, 3 September 2009 (UTC)

See below, I think we need to combine implications, such as recent gene-flow from NW Africa into SW Europe. You have now a collection of references on how best to deal with ambiguous implications (Schaffner 2004 for instance, Takahata 2001) and that the most effective way is to have evidence from a variety of sources such that one is not leaning to much on a single locus, such as Y.PB666 ^yap 16:11, 3 September 2009 (UTC)

On this I agree, but I've worked on what I know best today.--Andrew Lancaster (talk) 16:13, 3 September 2009 (UTC)

Getting things going

I have a couple of recommendations here, before we fix the name, let's first think how best to get the published materials into the article, particular stuff that was removed from the Genetic History of Europe. Since these names were drawn of usage on other pages, they are fine for now.

-Critical problems of the page. How material is introduced.

1. This articles topic is constrained by the following:

That ancient Europeans arrived primarily from Asia, not Africa, that to the best of our knowledge this first migration occurred 32 to 40 kya. And that these people had evolved since leaving africa (both with regard to selectable markers and with regard to neutral markers-neutral evolution)

That the evidence presented in this article pertains to gene-flow from Africa, indifference to previous gene-flow into SW Asia ~50,000 years previous, that is in a more recent time frame.'

2. That broad patterns of gene-flow, such as from the Middle East or from SSA via unknown paths or from the middle East need to be conditioned into a section of Alternative explanations, such that these are set aside from items that have more obvious connections. EV-13 might fall into this category.

3. The Lead should be a synopsis, the basic facts and refer to section leads, keep it short and free of opinion.

Describing information

1. Lets not make the mistake of Genetic History of Europe with long boring discussion of techniques, if people what to read on techniques they can read the various pages.

2. Contain information that pertain to single sources, for example, information about NW africa contribution should be presented in one section, if there is alot of information with Y or mtDNA than break it down into subsections, otherwise it should be combined. Figures, diagrams, maps are key to making a nice page.

3. Why is this information important, for example, what distinguished the EM-81 in northern Iberia from theories concerning Moorish migrations? Opinions are less important than style, this is an encyclopedia such that the reader needs to be coaxed to read.

4. Small sections at a time, work on each section so that when it is complete it is close to the final product, avoid opinion, even author opinion, each section should be validated with facts and conclusions from several sources.PB666 ^yap 16:11, 3 September 2009 (UTC)

I'll go through your comments...

• First whatever points you raise, I see no reason to stop thinking about what the real title should be. Focus has been a big problem with this article, and I think we can be pretty sure that delete debates have not stopped. I think it is fine to ask "what is the body of sourced material we are trying to put together" but this question can not be separated from the question of what it is that links these materials. We do not just want a mass of loosely relevant information.
• Intro. I lengthened the intro today and this may be one thing you are addressing. When I read your comment I went and snapped off what I added into a new section which lays out the basic timeframes involved. I stick by saying that not doing this will leave a gap where edit wars will put down roots.
• Your point about African routes going through the Middle East is an interesting one this article has to face. From a point of view of Europes genetic makeup is the disctinction being made between Africa and the Middle East perhaps even a little artificial? (Shock horror.)
• Long boring discussions. So far I think the article length is OK? What do you think?
• Are you arguing that it is better to break the main part of the article into regions than into types of DNA?

--Andrew Lancaster (talk) 16:26, 3 September 2009 (UTC)

No major complaints with the current content. My only concern is, as Andrew has pointed out, focus. Personally I don't mind detail and even prefer it, but the POV and OR police are always looking for excuses to slap fact tags. The more compact and direct the content is, the less opportunity is given for tagging.

The section "examples of gene flow" was originally "Defining African admixture". The reason was to address some of the controversies or misunderstandings that are currently prevalent concerning the origins of certain markers. Specifically that certain markers may have moved back and forth between Eurasia and Africa. Wapondaponda (talk) 07:12, 4 September 2009 (UTC)

E1

I am looking to add more information on Haplogroup E1 (Y-DNA). Though less frequent, it is an interesting haplogroup as it is found all over, including Italians, Albanians, Turks, Iberians, Canarians, Middle Easterners and North Africans. Haplogroup U6 also needs some mention as it has been detected in Iberia.

Wapondaponda (talk) 07:24, 4 September 2009 (UTC)

Elephant in the room

Based on the latest autosomal DNA evidence using the STRUCTURE program, Europeans don't have any African admixture to speak of:

• Europeans compared to Yoruba, Mandenka, S. African and Kenyan Bantu, San and Pygmy

• Same samples typed at over 600 times more autosomal SNPs

• More European samples compared to Yoruba

• Europeans compared to Sierra Leone and Tanzania

• Europeans compared to South African Bantu, Ethiopians and Afro-Caribbeans

• Europeans compared to virtually all Africans

So there's absolutely no justification for an entire article on this subject. It can only be a POV fork. ---- Small Victory (talk) 13:49, 4 September 2009 (UTC)

The thread was refactored, it was declared by NOR noticeboard to be original research. The NOR tag was also removed since the NOR here was removed.PB666 ^yap

Editing or deleting other people's comments is a big no-no. And nothing was "declared" by anyone. There's a discussion that's still ongoing, but no one has been able to demonstrate that citing a clearly explained chart is OR, because it obviously isn't. Besides, this is just the talk page. Deleting material from here that's under discussion takes information suppression to a whole new level. ---- Small Victory (talk) 12:06, 7 September 2009 (UTC)

Overtagging

According to the essay at WP:OVERTAGGING, no more than two or three tags are normally necessary. Furthermore reasons for adding a specific tag need to be articulated on the talk page. As I have not seen any reasons for some of the tags, I suggest removing some or all of the tags. If the details regarding any disputed content are expressed, then the tags can be readded. However, it should be noted that other means of resolving content issues should be attempted before adding tags. Wapondaponda (talk) 06:42, 5 September 2009 (UTC)

POV forking by SOPHIAN

Just fresh out of a month long block SOPHIAN has created Black African genetic contribution to the population of the European continent. Wapondaponda (talk) 06:53, 5 September 2009 (UTC)

Moorjani et al , new study

Abstract, is available. Confirms African ancestry in many southern Europeans, though percentages are not high, but they are nonetheless statistically significant, approx 2%. Furthermore the spanish admixture is estimated to have occurred 1700 years ago. Wapondaponda (talk) 12:44, 5 September 2009 (UTC)

We'll have to wait and see what these "methods that can infer admixture proportions in the absence of accurate ancestral populations" turn out to be, and if they have any validity. It sounds very dubious to me. Anyway, I don't know what you're getting so excited about. The study shows that admixture is extremely low and occurred in historical times, pretty much disproving everything you've ever claimed and confirming that this article is nothing but Afrocentric POV. ---- Small Victory (talk) 12:16, 7 September 2009 (UTC)

Wow. You even complain when Muntuwandi's African claims seem too unambitious. Anyone would think you are just looking for an argument.--Andrew Lancaster (talk) 12:27, 7 September 2009 (UTC)

The presentations are next month, don't know what the actual publication date is. Many of the authors are faculty in the Department of Genetics at Harvard. Reich who is already cited for Principal component analysis of genetic data is a coauthor. I don't know what methods they used, who knows may be even STRUCTURE. Don't know if I will even be agreeable to the results of the study. The abstract of Characterizing the history of sub-Saharan African gene flow into southern Europe states

Recent analyses of whole-genome SNP data sets have suggested a history of sub-Saharan African ancestral contribution into southern Europe but not in northern Europe, consistent with previous analyses based on the Y-chromosome and mitochondrial DNA. However, there has been no characterization of the proportion of African admixture in southern Europe, or of its date. Here we analyze data from ~450,000 autosomal SNPs in the Population Reference Sample, ~650,000 SNPs from the Human Genome Diversity Panel, and ~1.5 million SNPs from the HapMap Phase 3 Project, and studied patterns of correlation in allele frequencies across populations to confirm the evidence of African ancestry in many southern European populations but not in northern Europeans. Using methods that can infer admixture proportions in the absence of accurate ancestral populations, we estimated that the proportion of sub-Saharan African ancestry in Spain is 2.4 +/- 0.3%, in Tuscany 1.5 +/- 0.3%, and in Greece 1.9 +/- 0.7% (1 standard error). We also studied the decay of admixture linkage disequilibrium with genetic distance, which provided a preliminary estimate of the date of African gene flow into Spain of roughly 60 generations ago, or about 1,700 years ago assuming 28 years per generation. This date is consistent with the historically known movement of individuals of North African ancestry into Spain, although it is possible that this estimate also reflects a wider range of mixture times.

Firstly the authors attribute this sub-saharan influences to the movement of individuals of North African ancestry. This addresses the fact there has not been a clear cut distinction between North Africa and Sub-Saharan Africa, especially in prehistoric times. I don't know yet whether they address the migration of e1b1b lineages. Though the admixture levels are not high, they are still significant. In Spain, for example admixture of 2.4% was detected. With a current population of 46 million, 2.4 percent is equivalent to 1.1 million Africans, which is not a small number. Wapondaponda (talk) 14:06, 7 September 2009 (UTC)

STRUCTURE and admixture

According to SV's most recent posting on this talk page, SV has disputed this article based on results from the software program STRUCTURE. SV has argued that the presence of "discrete clusters" produced by STRUCTURE, means that no admixture exists between discrete clusters. Though STRUCTURE can be used to measure admixture, none of the studies cited by SV have published any data that directly concerns African admixture in Europe determined by STRUCTURE. As a result, the consensus from original research noticeboard has been that unpublished analysis of published facts constitutes original research. SV has been unsatisfied with this assessment and now it seems his whole argument centers on this software simulation.

I have been researching the apparent paradox as to why the authors of STRUCTURE say that it can be used to measure admixture but yet almost nobody has published a direct admixture analysis based on Structure. Despite the fact the STRUCTURE is used extensively, so far I have only seen one study, that used STRUCTURE to directly quantify admixture, and that is Auton et al for Mexicans. In the same study STRUCTURE was run on the global level and with K=5, five clusters which correspond to the major continents. A visual analysis of the chart indicates some discrete looking green and blue clusters, which SV has argued imply no admixture. However in the same study the authors have decided to address the "current debates in human population genetics regarding the extent of historical gene flow among Africa, Europe, and the Middle East". To do this, they do not use STRUCTURE, but instead use another method they call "haplotype sharing" which I think simply means counting the number of mutations shared by Yorubas and Europeans. It is interesting that they use STRUCTURE to measure admixture in Mexicans but not to measure Yoruba admixture in Europe. Furthermore they also perform a principal component analysis which revealed an African, Middle Eastern and European cluster, this runs contrary to the chart from STRUCTURE, that revealed separate discrete looking clusters of Europeans and Africans.

The simple reason for these apparent paradoxes, I believe is to do with how STRUCTURE is designed. There are two reasons, I believe, one is the assumption that the number of clusters, K is real and the second is the assumption of complete linkage equilibruim.

Value of K

The creators of STRUCTURE, Pritchard et al state:

here are also biological reasons to be careful interpreting K. The population model that we have adopted here is obviously an idealization. However we pointed out in our discussion of data set 3, clusters may not necessarily correspond to “real” populations.

This is illustrated with the study by Li et al and Li et al 2008 supplementary materials. They performed Structure using 650k SNPs on the global HGDP samples. When K=5, Druze have small amounts of African admixture, when K=6 this African ancestry disappears and the amount of South Asian ancestry increases in the Druze. When K=7 a new cluster, dark brown, which is prominent in middle eastern populations emerges. Interestingly not a single individual has 100 percent membership in this dark brown cluster. As can be seen, the value of K significantly changes the proportion of admixture. What value of K to use will depend on the specific study and any assumptions. Rosenberg 2003 and Auton 2009 used k=5, Li k= 7, and Tishkoff 2009 used k=14. Which K is correct is anyone's guess. Because of the subjective nature of K, it is understandable why STRUCTURE has not been used extensively for admixture mapping

Linkage assumption

The second assumption for Structure is that all loci within a cluster are in linkage equilibrium, Pritchard et al state that their model assumes that "loci are unlinked and at linkage equilibrium with one another within populations". Recent admixture introduces linkage disequilibrium into the receiving population and STRUCTURE will detect this and assign the markers in LD to the admixing population not the admixed population. In short, STRUCTURE is better equipped to measure recent admixture than more ancient admixture. This is because older admixture is thoroughly mixed in the population and as a result structure assigns it an indigenous status.

Summary

Basically STRUCTURE is a computer program that the phrase Garbage in garbage out applies. If all the assumptions are correct, then Structure will produce accurate data. But it seems this is not often the case because STRUCTURE isn't commonly used for admixture mapping. Wapondaponda (talk) 21:26, 6 September 2009 (UTC)

---

Will this never end? You're like a creationist trying desperately to discredit evolution, but failing miserably. A few replies:

• There has been no "consensus at the original research noticeboard". You wish.

• All STRUCTURE studies that conduct admixture analyses directly quantify admixture. And every study I've cited contains published data that directly concerns African admixture in Europe determined by STRUCTURE. Each includes African and European samples, and the authors make clear that the charts show continental ancestry and admixture proportions:

Rosenberg et al. 2005: We utilized the unsupervised clustering algorithm implemented in STRUCTURE [12,13] to group individuals into genetic clusters in such a way that each individual is given an estimated membership coefficient for each cluster, corresponding to the fraction of his or her genome inferred to have ancestry in the cluster. [...] Each individual is represented by a thin line partitioned into K colored segments that represent the individual's estimated membership fractions in K clusters.

Auton et al. 2009: In order to quantify patterns of population structure and admixture, we utilized STRUCTURE [Pritchard et al. 2000], a commonly used Bayesian clustering method. [...] Setting the number of clusters (K) to five revealed structure largely corresponding to continental regions (Figure 1A).

• The reason only Mexicans are discussed by Auton is because they're the only sampled population that has admixture. How many times am I going to have to explain the obvious to you? And how many times are you going to ignore it?

• Haplotype sharing is not a "method", and it isn't the same thing as direct admixture. Auton gives other explanations for it, including North African admixture in both SW Europeans and Sub-Saharan Africans, and simple genetic drift due to Iberia's role as a population refugium. But you know this very well because you're trying to suppress that information.

• There is no analysis that reveals an African, Middle Eastern and European cluster. That's just this plot viewed from a different angle (PCs 3 and 4 rather than PCs 1 and 2). And it's the first two PCs that Auton (and everyone who knows anything about PCA) says show continental ancestry and admixture that corresponds to the STRUCTURE charts:

The first two principal components (PCs) of the combined dataset separate individuals into clusters largely determined by geographic origin (Supplementary Figure S1A), which is consistent with a previous analysis of the HGDP dataset [Li et al. 2008].

We repeated the analysis using the 'supervised' STRUCTURE mode, having pre-assigned European and East Asian individuals to their respective populations. A K = 3, we found this method to give similar results to the unsupervised mode, with a European admixture component of 35.0% (standard deviation 16.8%) in Mexican individuals. The first two principal components of the same individuals demonstrates a similar pattern (Figure 1B), with Mexican individuals forming a distinct cluster between the European and East Asian Clusters in the first principal component. However, the second PC further differentiates the Mexican individuals from the East Asian individuals without substantially increasing the separation from Europeans.

• As you've noticed, increasing the value of K often reduces admixture. That's because STRUCTURE assigns individuals to clusters adding up to 100%, and if there aren't enough clusters to represent the true genetic structure, then you get false assignments. So there have to be at least as many clusters as continental groups studied. For example, in Wilson's analysis, no Native Americans were tested, so 4 clusters were enough; this wasn't the case in Rosenberg's analysis, where at least 5 clusters were needed to differentiate the Oceanians, who showed exaggerated Asian, European and African admixture prior to that. Therefore, STRUCTURE may overestimate African admixture in Europeans, but it won't underestimate it, despite what you're hoping. ---- Small Victory (talk) 12:51, 7 September 2009 (UTC)

There was a consensus at the noticeboard, a very stark one. And you are not understanding what people are saying to you, because you do not want to, and because as usual you have lost track of your own tortuous efforts to argue with people about every little thing. You have your hands over your ears. Anyway, this article is not about STRUCTURE, and it does not need all the OR you want to put in because if all you want to argue is that analyses tend to show admixture to be small, then who is arguing with you? You don't need a debate about STRUCTURE for that. Everyone seems to agree on that? Indeed you just posted a sarcastic message on this talkpage to point our how un-impressively Afrocentric Muntuwandi's recent edits were. You apparently have no point worth making.--Andrew Lancaster (talk) 13:02, 7 September 2009 (UTC)

Well according to Auton et al, they state the STRUCTURE, underestimated admixture in their study,

To investigate the level of admixture in the Mexican population, we combined the Mexican samples with a sample of European and East Asian populations. Using structure with K = 3 we estimated an average of 32.5% European ancestry in Mexican individuals which is lower than some previous estimates based on microsatellite or "ancestry informative" markers

Increasing the value of K does not necessarily reduce admixture, because new clusters emerge. As in the case of Li et al 2008, ancestry in the south Asian and middle eastern cluster increased from k=5 to k=7, whereas African and European ancestry reduced.

It seems that SV has adopted a one-dimensional approach based only a STRUCTURE. It seems the only reason SV clings to STRUCTURE, is because it is the only method that you believe supports your position. Better to look at the bigger picture, which includes, y-DNA, mtDNA, HLA, PCA, genetic distances, archeology, anthropology and linguistics, which are all quite consistent with each other. Wapondaponda (talk) 14:28, 7 September 2009 (UTC)

We've already been over why Auton's results differed from previous estimates:

However, it should be noted that the variability between individuals is high, and that the Mexican samples in our study originate from a single location (Guadalajara).

That has nothing to do with what I'm talking about. Obviously, when new clusters emerge, ancestry in those clusters increases. It's the ancestry in the preexisting clusters that decreases, as misclassified individuals get assigned to the correct new clusters. Do you understand any of this stuff?

And I don't "cling" to STRUCTURE. It happens to be the latest, most widely used method of genetic testing, and the only one designed specifically to quantify admixture, which is what this article is supposed to be about. Rather, it's you who's clinging to outdated, deprecated and irrelevant methods because you believe they support your position, and you know that STRUCTURE analysis doesn't. ---- Small Victory (talk) 11:55, 8 September 2009 (UTC)

Going through the current article - "First settlers"

The majority of inherited genes found in modern Eurasians were carried by a small number of individuals that left a constrained area of East Africa more than 70,000 years ago.[2][3] Currently accepted models suggest migrations from Arabia into SW and South Asia. Over time some peoples migrated northward and westward into South Eastern and Eastern Europe (See Peştera cu Oase) after 45,000 years ago. At this time Neanderthals lived in Europe and areas of Southwest and Central Asia. While there is still a possibility that some Neanderthal genes entered modern humans in the past, most of the genetic makeup of Europeans can be traced to the early migrations Out of Africa.

— Version right now

I see a lot of problems with this.

• First settlers as a title leaves a lot of explaining to do. Who decided who the first settlers are? In other words why are Cro Magnon the first and not the Neanderthals or other archaic humans? In order to help readers understand this the distinctions between archaic and anatomically modern needed to be mentioned but this has been removed.
• First sentence implies consensus in the literature which does not exist. Many authors think the genetic landscape of Europe results largely from things happening when farming was introduced or after. The sources given are two Tishkoff articles, one from 1996, and one about a different subject.
• The third sentence then goes on to explicitly imply there is consensus opposite to what the first sentence demanded!
• The second sentence implies that there is consensus that the Sinai was unimportant. Why?
• Last sentence does not define "the early migrations Out of Africa" which people who know the subject will know is unrealistic, and which will mislead people who do not know the subject.

Several of these problems keep getting inserted into the text and have been pointed out before. (Or perhaps it is more correct to say that important information keeps getting removed leaving big holes.--Andrew Lancaster (talk) 05:49, 7 September 2009 (UTC)

What text do you suggest. Wapondaponda (talk) 06:16, 7 September 2009 (UTC)

It might be easier to just start editing, because we are not in an edit war, but I see the first and second sentences as basically needing to go (exact migration routes between Africa and Asia not relevant anyway) ; and then some extra information has to be inserted to make some points more clear.--Andrew Lancaster (talk) 07:09, 7 September 2009 (UTC)

Looking ahead at questions that might be raised

For the time being, this subject now has a chance to settle into an article and justify its existence. (Small Victory and SOPHIAN managed to get themselves blocked for those who don't realize.) We do not know how it will look after a month or so, but it is possible that the subject matter will substantially overlap with Genetic History of Europe. I thought it worth already stating that my thinking about the questions this raises would be like this:

• If the content could be merged in Genetic History of Europe with creating big diversions, then I think everyone should be happy that we had a chance to polish up this material and merge it back into a presumably better article? The extra discipline of having to work on one theme might even lead to this article containing preferable materials for some sections, and indeed I suggest that anyone editing either article should look to the other every now and then for ideas.
• If it is not possible to insert whatever this article becomes without substantially messing up the Genetic History of Europe article with diversions etc, then I guess this here article (with a better name?!) should stay.

Is this also how others see it?--Andrew Lancaster (talk) 10:10, 7 September 2009 (UTC)

Eventually some content will have to be merged back into Genetic history of Europe. But it should follow the format of WP:Summary style. In other words, we should have a summary of this article in Genetic history of Europe, and we can maintain this article. The main reason for maintaining this article has to do with space. In general the content in GHOE should be proportional to the proportion of admixture, so we cannot have too much information about African admixture on GHOE. But there are no limits to how much information we can place in this article, as long as it is directly relevant and notable. Wapondaponda (talk) 10:32, 7 September 2009 (UTC)

In terms of parts of these two articles being merged, this can always happen and indeed I did some edits to both today, based on comparing them, and I can see more. But what if we come up with an article which is essentially covering subjects which are ALL covered in the other article, and could be merged back in without making the other article worse in any way including length? What I guess I am saying is that if this does happen I see no problem merging it back in.--Andrew Lancaster (talk) 11:07, 7 September 2009 (UTC)