More woeful misunderstandings of DNAPrint tests

Posted by JW Holliday on Monday, 01 August 2005 15:40.

The latest in a series of critiques of the DNAPrint tests is at John Hawkes’ blog.  This, as usual, is based on certain, not unpredictable misunderstandings.

As I’ve commented already on these tests, I’ll just give more of an outline format for a response.

1. While Hawks does mention the issue of statistical significance, he doesn’t follow through on it to completion.  If a low level of affiliation (eg, “Native American” in Greeks) is below the level of statistical significance and the “confidence intervals” overlap zero, then that result is statistically equivalent to zero.  Given that tests using continuously distributed alleles will have a low error rate, and given that levels of statistical significance are given on the company’s website, these low level results should not be taken too seriously.

The company is undertaking putting together a more advanced version of the tests that may in part clear up what is going on with these very low affiliation results.  However, it is quite clear that making a big deal, for example, out of a 4% Native American result in a Greek is a waste of time.  It is not significant; the company admits that, so why make an issue of it?  That some people misunderstand and/or misuse the test does not logically invalidate the worth of the test itself.

2. A serious misunderstanding – and one that DNAP itself may have some responsibility for because their explanations of the tests are rather poor – is the idea that DNAPrint’s categories somehow represent the specific populations they are named after in a direct fashion, and, in so doing, the categories represent direct descent from these “pure populations” – which Hawks sees as a serious flaw in the assumptions built into the tests.

The problem is that the categories obviously do not directly represent any modern populations since, for example, real-life South Asians and Middle Easterners are NOT 100% “South Asian” or “Middle Eastern” when tested with DNAP.  Furthermore, DNAP admits, for example, that they really are not sure what their Euro 1.0 categories exactly represent; one cannot say that DNAPrint’s categories represent ancient “pure populations”, either.  The straw man of “racial purity” is irrelevant here.  There needs to be no “pure populations” as an assumption for any of these tests.

Important - What the categories represent are those sets of gene frequencies that characterize the predominant distinctive ancestral genetic component of particular population groups.

That is much different than saying the category represents the group itself.  It may well be, for example, that, even in ancient times, the populations of the Middle East were never “racially pure.”  Nevertheless, these populations share a particular predominant genetic component that distinguishes them from other major population groups, and this component can be determined and roughly quantified via the DNAPrint tests.  Thus, the assumption built into the test is that different populations have these distinctive genetic components - not that the populations themselves, including the parental populations from which the gene frequencies were determined, are “pure” or not.  In other words, the tests are pulling out from these populations that predominant component that distinguishes the population from others; that same ancestry may well be a minor component of the ancestry of another group.

3. Therefore, even if there is statistically significant “Native American” in a Greek, the test is not saying that the Greek in question is descended from Native Americans!  Rather, that the Greek contains, for whatever reason, a minor genetic affiliation with gene frequencies that represent what is the predominant distinctive portion of Native American SNP variation.  DNAPrint openly admits that they cannot distinguish how and why these gene frequencies are present:

Thus, the DNAPrint tests are really not genealogy tests per se, nor are they tests of absolute racial ancestry.  One cannot take a set of DNAPrint results as a literal reading of one’s exact ancestry.  One cannot say, hey, 35% of my ancestors were Northwest Europeans, 15% were South-Eastern Europeans, 10% were Middle Easterners, 10% were Native Americans, 10% were Africans and 20% were East Asians.

Instead the tests are relative and comparative, they allow you to compare how individuals and groups compare to other individuals and groups with respect to their relative genetic affiliations to the predominant distinctive gene frequency patterns of various major population groups.  Different ethnic groups tend to have distinctive patterns of these affiliations.  So that, one does not say that a South Asian is “100% South Asian”, but that South Asians tend to be characterized by a set of affiliations including a predominant “South Asian” component, usually with a strong “East Asian” minority affiliation, etc.

Whether these affiliations can be correlated to actual recent (or ancient) admixture depends on context.  A person who tests as “60% Native American” is undoubtedly significantly derived from recent Native American ancestors.  Some white person who is “12% Native American” – that depends.  If they are a Greek from Athens, obviously they are not having direct ancestry from actual Native Americans.  If they are white Americans whose ancestors came to the New World in the 18th century, and may have admixed with Native Americans, that is a stronger possibility.

Updated future tests may be able to “time” the “admixture” event based on the level of clustering of particular alleles, but for now, it is not possible.

The results must be viewed in the context of what is known of the person’s background; in this case, when used in a comparative sense, focusing on statistically significant results, then they are very informative.

4. I don’t quite understand Hawks’ problem with the South Asian data.  South Asia is at the geographical/racial crossroads between the Caucasian and Mongoloid worlds.  Furthermore, a variety of genetic studies (which I believe I’ve already mentioned on this blog) demonstrate the unique, mixed nature of South Asian populations and, in particular, East Asian influences (HLA studies, as well as the Ray et al Alu work).  Looking at populations of northern and eastern India, Nepal, Bangladesh, Burma etc, I fail to see why it is so impossible that South Asians are a Caucasian/Mongoloid mix, with the former of course predominating, while certain SE Asian populations may contain a certain Caucasian influx from South Asians.  I fail to see as well why the people of Hawaii would not test out as predominantly in the East Asian/Pacific Islander category, why East Africans are not a genetic mix of African/Caucasian, and I thought as well that the Asian influences in Madagascar are well known, along with the obvious African components.  Cannot the same criticisms be made of other studies, such as those from the Rosenberg and Risch labs?  Even if a limited number of population groups are used to look at variation, for most populations, useful information can be obtained.

In summary, I really don’t see major problems with the tests, when properly interpreted.  As well, newer tests are being devised.  Just because the tests may be misused and the data twisted to fit political objectives, this does not invalidate the tests themselves.

I see the major points as being that the tests are relative and comparative, rather than absolute, and that the names given to genetic ancestral categories do not exactly correlate with the specific ethnic/racial groups they are named after.  For example, some people seem to be disturbed that tested Irish were found to have a certain low percentage of “Middle Eastern” and “South Asian” affiliations.

Assuming these are above statistical significance, then, how can that be?? – after all, the Irish are NW Europeans with no history of being “invaded by Arabs or South Asians.”  But, that’s the point, the data are NOT saying that there must be direct descent from those particular groups.  Instead what the data simply show is that the predominant distinctive genetic (ancestral) components in Middle Easterners and South Asians are also found as minor components in the gene-pool of the Irish.  Why is not known, but it does not logically require direct descent from one group into another - simply a sharing of allele frequencies.

That is, admittedly, not genealogy.  But as I’ve said, these are not really genealogy tests.  Rather, they are tests of relative racial/genetic affiliation.  The data, and in particular the minority affiliations, are NOT randomly scattered among individuals and groups - as they would be if they had no relevance. They are structured in a manner consistent with broad racial clines and historical trends (eg, elevated Middle Eastern in Ashkenazim) and they correlate with phenotype, too.



Posted by Geoff Beck on Mon, 01 Aug 2005 17:22 | #

I am not trained in any biological sciences, yet I have experience with professors of Anthropology and Sociology, and all of them are fervent race deniers.

Many will go hysterical lengths to defend their thin theories. Anyone who has experienced, as I have, the politicized courses in social sciences must discount any pronouncements by that crowd.

Today Anthropologist and Sociologists have no purpose other than to attack Western civilization, and deny the White his identity. Period.


Posted by Steve Sailer on Mon, 01 Aug 2005 18:45 | #

Good write-up.

The distinction between African genes and everybody else’s is clearer than between Eurasian genes.

I commented on my blog awhile ago:

AncestrybyDNA’s new Euro-DNA 1.0 genetic test—Genealogy is a huge hobby and genetic testing services have emerged to sell to people interested in their ancestral backgrounds. Most have focused on the easier to test Y-chromosome (direct male line) and mitochondrial (direct female line) tests, but those only tell you so much. In contrast, AncestrybyDNA offers tests on your autosomal or “pan-genome” DNA which give you a statistical picture of your whole family tree.

They’ve now introduced a test for Caucasians that can tell you your admixture as broken down into four groups:

Northern European subgroup (NOR)
Southeastern European (Mediterranean) subgroup (MED)
Middle Eastern subgroup (MIDEAS)
South Asian subgroup (SA)

Their paper introducing the service gives lots of interesting details on the genetic background of Europeans and white Americans. In general, Europeans appear to be the most homogenized of all the continental-scale racial groups, so I suspect that customers are typically going to be underwhelmed by the findings: E.g., “You’re 50% NOR, 35% MED, 10% MIDEAS, and 5% SA.” That’s probably not going to satisfy somebody who puts a lot of emphasis on his ethnic identity as a Walloon or Slovak or whatever. But, the write-up is most interesting.

The entire website is loaded with neat stuff, but it’s important to keep in mind that this autosomal DNA analysis is still a long way from being 100% accurate. For example, here is one family where the mother was measured as 22% black and the father as 0% black. The test reported that the three children were 15%, 2%, and 6% black by ancestry—not bad, but not great either. There’s a certain amount of statistical noise in the results, so the website offers a lot of caveats—for example, a lot of Jews who know all four of their grandparents immigrated from Poland take the test and show up as a few % Native American. So, don’t put much weight on small numbers.


Posted by JW Holliday on Tue, 02 Aug 2005 08:54 | #

Thanks, Steve.

I think that DNAPrint itself, and Shriver, can be legitimately critiqued for some of the facile comments they’ve made in the past, for example, the “revisionist history” theories for some of the data, as well as pushing an “we’re all admixed” agenda, stressing the “admixture in white Americans” (eg, this % of white Americans have admixture from “X”)in public statements *without* clearly stating that for many (most?) of those people, the “admixture” is below the levels of statistical significance, and is so indistinguishable from experimental error.

But, with respect to the latest versions of the test (although the first version was not powerful enough, with only 71 markers and should not have been released), I see nothing at all fundamentally wrong with the test itself, nor with the data produced, as long as these data are conservatively interpreted.

I think a lot of people are going ballistic about these minority affilation percentages (eg, Native American in Greeks; South Asian in Irish) without reflecting that they do not necessarily imply any direct descent from Native Americans or South Asians.  These are shared gene frequencies, not shared ancestry.  The shared gene frequencies may be due to shared ancestry, or they may be due to something else entirely.

In any case, I take a “glass half full” approach.  DNAPrint, despite any misgivings, needs to be encouraged to continue developing the ancestral products.  If they would drop that in favor of a pharmacogenetics-only approach, that would be highly disappointing.


Posted by Fred Scrooby on Sat, 06 Aug 2005 00:04 | #

John Hawks posts a <u>rejoinder</u> to this log entry at his <u>blog</u>.  The first two thirds seem devoted to debunking “pure races,” which is needless since no one at traffics in the notion of “pure races” but only in the notion of “races pure and simple” (in other words, put “pure” after the noun and you’ve got our view).  For the Nth time we here don’t have a notion of “pure race.”  The notion of “race”—just plain ordinary “race” thank you very much, all by its little self, without the “pure” strawman in there anywhere—serves truth perfectly.   

Hawks goes on to ask,

“But in what sense, exactly, does it mean anything to say that East Africans are a mix of Africans and Caucasians? Or that South Asians are a mix of Caucasians and Mongoloids? Indeed the question presupposes a hypothesis of racial history that genes do not support. [...] Do the allele frequencies provide any evidence of this ancient mixture of populations? What, exactly, were these ancient ‘Caucasians’ and ‘Mongoloids’ if they were not the ‘pure races’ the test supposedly is not assuming existed?”

Hawks acts as if it would make sense to interpret the genetic data as reflecting an ancient Subcon population whose genotype combined what are now considered Mongoloid and Caucasoid allele frequencies into one, then split up, giving rise to the Caucasoid and Mongoloid races of today.  Apart from being thermodynamically backwards and therefore extremely unlikely to say the least, that violates Occam’s razor, big time.   

“This may sound like a word game.” 

No, it just sounds like someone who’s laboring mightily to convince himself he needs to reject Occam’s razor in obedience to the demands of politics, in favor of Ptolemaic-style epicycles-on-epicycles as the explanation of things that really have a much more clear explanation, one however which politics forbids.  That’s what it sounds like.

“Is South Asia ‘at the geographical/racial crossroads’? Why ‘of course’ should Caucasian genes have predominated if this were true?”

Who said if they mixed they had to come out exactly fifty-fifty?  Whoever said that?  Who knows what forces might have made one predominate?  What does that matter?  How does the fact they didn’t come out 50-50 cast doubt on the theory that they mixed?  I mean, when the Missouri River enters the Mississippi does the water from each have to occupy exactly 50% of the river bed all the way to the Gulf of Mexico?  If one predominates does that cast doubt on the notion that they mixed?   

“Then why are we talking about ‘distinctive genetic components’ at all? Why don’t we just talk about allele frequencies?” 

Fine, let’s talk about allele frequencies.  What’s the problem?  Who’s insisting on “distinctive genetic components,” anyway?  No one I know of.

“I think the comparisons would mislead any non-knowledgeable person to believe that he actually had genetic input from different races, and that these races are meaningful categories representing ancient human groups.”

You know what this sounds like?  It sounds like Hawks wants to be PC and deny races but doesn’t want to make himself a laughing stock by denying those of today’s world, so is going back to ancient times and trying to cast doubt on the existence of races at that time.  There’s every reason to suppose races existed then as well and were as differentiated as those of today if not more so.

“to the extent that a person goes to the trouble to understand detailed deviations from the racial categores, such as the high ‘Native American affiliation’ of Greeks, he is learning biological nonsense.”

No one here disputes that.

Wasn’t there anything in Hawks’ polemic that made perfect sense and also seemed to need saying?  Yes.  This:

“Who is misusing the test? For one, sociology professors trying to make their students shed their ‘traditional’ notions of race.”

I like John Hawks and read his blog regularly.  But clear talk from him on race is scarcer than hen’s teeth ... In fact, I’ve never seen it.  What’s he afraid of?  Or, what can’t he see?


Posted by JW Holliday on Sat, 06 Aug 2005 16:16 | #

Fred, thank you very much indeed for drawing our attention to John Hawks’ riposte.  Your are correct that we are NOT here talking about absolute racial purity, but rather, relative genetic homogeneity. Hawkes seems to live in a dream world in which there is no difference, racially speaking, between someone from Northwest Europe and an admixed person from Puerto Rico or Central Asia. There are no “pure races” you see, and the fact that the Irishman is descended from a more restricted set of lineages in evolutionary time means nothing. It’s not “politically correct”, you understand.

What nonsense. The categories used for the parental populations in ABD are of groups that represent the extremes of clines of gene frequencies, hence the archetypes of genetic races (gosh - non-PC!).

Population genetics studies can and DO discern the relative genetic distances and similarities of different groups, and determine which groups cluster and which do not. Northwest Europeans, Northeast Asians, West/Central Africans, and Central Amerinds represent populations that are particularly distinct among worldwide populations, while populations that are historically known to be admixed are less so distinct. Yes, that “bad” word distinct again.  Distinctive gene frequencies are what needs to be determined; since the majority of human genetic variation is randomly distributed at the population level, it is obvious that a test that wishes to determine relative genetic affiliations needs to concentrate on that portion of the genome that is population distinctive. Any person with a triple-digit IQ can understand that. Perhaps certain ‘academics’ cannot.

As far as I know, the DNAP site says that the Euro 1.0 categories were determined empirically, based on clustering. If PC ‘scientists’ have a problem with empirical science, let them curl up in bed with the work of Gould and Lewontin and feel nice and cosy with such scientific obscurantism.

You know, human biodiversity developed by the differentiation of populations: African vs. Eurasian; Caucasian vs. Asian; and further subdivisions.  At each level, members of the resultant population groups share a greater number of comment recent ancestors than with members of other groups.  This in turn results in differences in distinctive gene frequencies.  Or perhaps, does Hawks deny the differentiation of populations, the clustering of groups, and that Swedes, Nigerians, and Koreans differ genetically.


Posted by JW Holliday on Sat, 06 Aug 2005 16:19 | #

Furthermore ...

As regards the worth of the tests, I see it as extremely useful to see how different groups and individuals compare with their distinctive genetic components. I do understand that for so-called ‘scientists’ who are “PC”, information that can lead to an understanding that people DO differ in a racial/genetic sense should be bashed. If people are too stupid to understand the tests, that is their problem. I agree DNAP needs to do a much better job of explaining the tests.

With respect to whether the results can be used to determine actual ancestry, any literate person who actually read my post can see that I stated it depends on context. I assume that to Hawks, the fact that a Mexican may test as 50% European and 50% Amerindian is ‘meaningless.’ God forbid anyone gets the idea that the Mexican genepool is a combinaion of Spanish and Amerindian, with a touch of African. How can that be!  By golly, there are no “pure races”, so let’s pretend that the genetic ancestry of Mexicans is unknowable and untestable, with respect to gene assays. And the same goes for any other group.

Another thing that the Hawskites cannot answer is why there is a correlation between DNAP data for different populations and what is known about these from population genetics and history. Not only the South Asians, but the clines of NW and SE European ancestry in the DNAP test match very well component 1 identified by Cavalli-Sforza, as probably represents paleolithic/neolithic ancestries. There are gradients of ‘Middle Eastern’ that make sense, and “it just so happens” that a group of Ashenazim have an elevated Middle Eastern component. Yes, indeed, the idea that DNAP can detect a ME component in Ashkenazim, which “just happens” to match NRY studies as well as history - that is all “bad” and “wrong” and “meaningless.”  As is the DNAwitness website that shows that the genetic components found in admixed individuals of Filipino or Mexican ethnic ancestry “just happens to” match their phenotype, AND what is known of the population histories of those nations.

The FACTS and the EVIDENCE support the validity of the DNAP tests. The sort of equivocating nonsense that spews forth from academia is just so much hot air. If the tests are “bad”, then why do they work?


Posted by JW Holliday on Sat, 06 Aug 2005 16:21 | #

And finally ...

With respect to comments made by Hawks concerning groups such as South Asians, East Africans, and the natives of Madagascar, I submit the following links.


East Africa:-


Third party readers can come to their own conclusions about Hawks and his competence in population genetics and the histories of varying population groups.  I myself cannot take this fellow seriously.

Who is more correct about South Asians – DNAPrint or Hawks?  Which view is more consistent with the literature?


Posted by Fred Scrooby on Sat, 06 Aug 2005 17:34 | #

Reading Hawks’ log entry one sees one of the lines of attack that will be mounted against genetic race testing (if not already mounted—I don’t keep current with this literature):  they’re going to try to call DNA-race-testing circular because it’s based on the traditional definitions of the various races from physical anthropology (what Hawks calls the Linnean races) which they’re going to condemn as meaningless, in typical race-denying fashion.  They’re going to try to get away with this in spite of Cavalli-Sforza’s work demonstrating independent genetic confirmation of the traditional broad “continental” racial categories, because the DNA testing companies have taken as their “DNA baseline profiles of each racial and sub-racial group” the DNA profiles of individuals they’ve chosen on the basis of how well they satisfy the racial and/or sub-racial criteria of traditional physical anthropology (combined with the known social, political, and demographic histories of the places where the individuals live, and supportive genealogical information about their family histories).  The other side are going to say that classifications based ultimately on traditional “Linnean race” categories are meaningless because those categories are meaningless. 

There’s only one way to get around this, and that’s to fully rehabilitate the traditional physical-anthropological “Linnean” racial categories.  Not that they need rehabilitating, as they’ve never been even remotely legitimately invalidated.  But the point is, there are lots of people out there who think nothing is clear until genetics makes it clear.  That’s wrong.  The traditional racial and subracial categories are valid—or, let’s put it this way:  if they aren’t, then neither is any other classification whatsoever:  birds, lizards, humans, monkeys, insects, trees, flowers, snakes, canines, felines, bovines, and so on.  If races are drummed out of traditional Linnean classification, everything else goes with them—sorry, leftists and PCers!  They are as valid as everything else that’s ever been identified and classified.


Posted by JW Holliday on Sun, 07 Aug 2005 16:53 | #

Hawks mocks the “low, low price” of the DNAP test; I wonder what the “low, low price” of tuition at the University of Wisconsin/Madison is.  More to the point, are anthropology students there getting their money’s worth?

Look at this guy’s purported specialties and interests and then compare that to the obfuscatory nonsense he has written.  Pay close attention to his comments about South Asians, East Africans, etc. compared to what has been published in the peer-reviewed literature.

This is astounding.  And disturbing.  While I often disagree with Dienekes, it seems to me that the non-specialist Dienekes Pontikos knows a lot more about population genetics and anthropology than does Hawks (compare their blogs and make up your own mind about it).  In my opinion, students about to spend their money (remember ... they are paying for those “low, low prices” at U of W/M) on Hawks’ anthropology course would be better advised to read Dienekes’ blog and scan through Medline on their own.

Hey ... anthropology with Hawks - what me worry?

Yes, students…worry.


Posted by Fred Scrooby on Sun, 07 Aug 2005 20:01 | #

He generally shows impressive knowledge and excellent sense but gives the impression he can’t get himself to deal clearly with race questions.  On those topics he dances around the issue being vague, never coming down squarely with an unequivocal affirmation such as “yes, race damn well does exist,” and even seems to flirt with the opposite claim, that the existence of race isn’t adequately demonstrated (while never unequivocally taking that position either).  He’s either afraid or he’s not thinking straight on race.  I could be wrong, but that’s the way he comes across to me.


Posted by Fred Scrooby on Sun, 07 Aug 2005 21:47 | #

Hawks <u>here</u> discusses the potential implications, mainly medical ones, of a cheap—a thousand bucks a pop—complete genome sequencing for an individual.  He predicts they’ll become commonplace for medical purposes and I agree, if the price is gotten down to that level—which surely is only a matter of time.  After a long, interesting discussion he finishes by saying young adults won’t be interested in the info it contains or the doctor’s (or computer’s) recommendations for each individual’s optimal health maintenance based on his exact genome, since college students and other young people view health problems expected to start cropping up decades later as too remote from their everyday concerns.  He may be right about that too.  But there’s one thing I predict many a young person will be interested in:  his individual racial heritage as reflected in his genome and also that of the person he’s thinking of marrying.  Take someone like Razib—you can be certain he’d consider the racial indices of any prospective mate’s genome if the genome were available and there were any question about her racial heritage.  Many are like him out there.  U.S. Jews have already greatly reduced the incidence of Tay-Sachs disease by this method:  getting tested, then two carriers being careful not to procreate together.  Tay-Sachs disease will finally be eliminated from the Jewish sub-population in this way but not the heterozygotes, since a person who’s not a carrier can marry one of those.  But the heterozygotes may eventually diminish too, since among those negative for the trait there will be some who won’t want to procreate with a carrier, so the latter will over time dwindle among a certain subset of the population. 

The same thing will happen in the general population, especially, I predict, the white-Euro one, in regard to certain genomic racial indices:  some won’t want to mate with a “carrier” of certain indices and these will eventually be bred out of certain sub-popuations—or rather, certain new sub-populations will form which will be lacking in certain racial genomic indices. 

I predict a class of people will emerge who’ll pay attention in this way to racial characteristics of prospective mates as reflected in their genomes.  No matter what happens on the immigration front, therefore, races will continue in existence in this way, because that will be important to some.

By the way, for any who were wondering about Hawks’ PC credentials, they’re on display in this entry’s second sentence, where he uses “her” as the generic personal pronoun instead of the standard “his”: 

“I suppose $1000 would be inexpensive enough that anyone with a persuasive medical need could get her genome:”

Mystery partly solved ...


Posted by manju on Mon, 08 Aug 2005 12:29 | #

Instead of looking at South Asia/Indian sub-continent an admixture between European and East Asian, it should be considered as a secondary repository of Homo Sapiens(Primary obviously being Africa). That will ease so many complications.


Posted by Fred Scrooby on Mon, 08 Aug 2005 14:32 | #

“Instead of looking at South Asia/Indian subcontinent as a mixture of European and East Asian, it should be considered a secondary repository of Homo Sapiens (Primary obviously being Africa). That will ease so many complications.”  (—Manju)

Sorry—Manju’s point escapes me here.  Are you talking about the multiregional theory, Manju, with the Subcontinent as one of the places of independent evolution of homo sapiens?

For the record, the passage below is the gobbledeygook that prompted comments here on the subject of the Subcons being a mixture of Caucasoids and Mongoloids (notice, by the way, the coy, irritating flirtation with race-denial in the last sentence):

<u>But in what sense</u>, exactly, does it mean anything to say that East Africans are a mix of Africans and Caucasians? Or that South Asians are a mix of Caucasians and Mongoloids? Indeed the question presupposes a hypothesis of racial history that genes do not support. Is South Asia ‘at the geographical/racial crossroads’? Why ‘of course’ should Caucasian genes have predominated if this were true? Do the allele frequencies provide any evidence of this ancient mixture of populations? What, exactly, were these ancient ‘Caucasians’ and ‘Mongoloids’ if they were not the ‘pure races’ the test supposedly is not assuming existed?  This may sound like a word game. But these words have a specific meaning and history. It seems to me much clearer and more accurate to say that some alleles have geographic distributions that include Greeks and Native Americans than to say that a Greek has a ‘Native American affiliation.’ The former has the advantage of being true, while the latter is quite obviously false—unless you are a geneticist classifying people by race.”


Posted by JW Holliday on Mon, 08 Aug 2005 17:28 | #


The issue is not how we “should” look at South Asia, it is not a matter of personal choice!  It is…what are the genetic facts that we can get from population genetics. 

First off, we can safely ignore Hawks’ strawman representation of the DNAP data as saying that South Asians are “part European and part East Asian.”  In fact, truth is best served by ignoring anything that obfuscator has to say.  Let’s look at what the DNSP data actually says.

It says that the predominant ancestry among (tested) South Asians is an Indo-European/Caucasian strain that
they term “South Asian.”  In other words, the predominant genetic strain in South Asians is a particular type of Caucasian that is differentiated from other types - Europeans, Middle Easterners, etc.

The data also say that a secondary minority affilation of South Asians is “East Asian”, that is, what is the predominant ancestry of the Mongolid peoples.  South Asians also have some genetic affiliation to “Native Americans” (most likely Central Asian), as well as to Africans and to West Eurasian Caucasians.

Now, if one looks at some of the literature, including the links I presented in this post, there is a reasonable correlation between the mass of the peer-reviewed literature and DNAP data.  Obviously, there are many questions.  Obviously, many things need to be answered.  Obviously, new data may change the current picture.  All of that is acknowledged.

However..that does not change the fact that South Asians are not some mysterious unknown genetic entity completely disconnected to others.  I see no strong evidence that South Asians represent one of the orginal branches of humanity, but rather arose after the differentiation of Eurasians vs. Africans.  The data do support some East Asian-type genetic influx into South Asia.  I await further data to confirm or refute this, and I understand that there is no uniformity even today about this - I am merely summarizing what the bulk of the literature points to, there may well be some counter-views.

But, there is really no strong and convincing evidence that the DNAP data for South Asians is “wrong”, within the confines of the literature.

Furthermore, it is plain wrong to characterize the DNAP data as saying that South Asians are part “European” and part “East Asian” (with European being defined in its strict sense), given the Euro 1.0 data that DNAP has on its website.


Posted by JW Holliday on Mon, 08 Aug 2005 17:36 | #

Some comments with respect to Dienekes’ post regarding Hawks and DNAP:

1. I’ll say it again and again, the DNAP tests to not per se measure “admixture.”  The tests measure relative genetic affilation.  Differences in affilation may be due to admixture or they may due to other reasons.  This has been discussed here already in depth.

2. Dienekes’ labelling of certain DNAP results as “problems” is, in my opinion, not justified without further evidence.  Given that “Middle Eastern” in the Irish does NOT imply any sort of direct admixture of Middle Easterners (at least, as they exist today) into the Irish genepool, how it is known a priori that this is a problem?  The data mean, ultimately, that what is the predominant genetic component in Middle Easterners is, for whatever reason, a minor genetic component of the Irish.  Given a common “recent” ancestry for all Caucasians after the split from the Asians, it is not so surprising that such overlap may exist.  The meaning of it remains to be determined. The same goes with “Native American” in Greeks or “South Asian” in Iberians.  The only “problem” is taking the labels literally. Furthermore, in many cases, the sample sizes are too small to say with certainty that there really is a problem.  Who were the “Iberians” tested?  A family of gypsies?  Who knows?

One must remember the tests are relative and comparative.  It is not the presence of ME in Irish that is so important, but why this component is larger or lesser in the Irish than in other Caucasian groups.

3. The DNAP methodology does not simply assign a MLE; the algorithm determines the probability of every combination of ancestries and produces a probability curve.  The MLE is simply the single most likely value.  DNAP provides confidence intervals with their data.

4. Dienekes’ comments about South Asians disregard the Euro 1.0 test.  Yes, indeed South Asians have a South Asian unique component as determined by mtDNA and NRY.  BUT…the Euro 1.0 test includes this component as a separate category, termed “South Asian.”  Thus, one does not simply look at ABD 2.5 and say that South Asians are composed of “4 categories.”  Actually, there are a total of 7 categories, one of which is a specifically South Asian component.  And, there is evidence for West Eurasian and East Asian components to the South Asian genepool.

5. With respect to chimps - this is why, for example, DNAP has strict requirements as to who is eligible for Euro 1.0.  Their algorithm looks at the ABD2.5 data to determine what genetic structures are such that the data would be meaningful.  A 100% African taking Euro 1.0 would be ‘forced’ into categories - but DNAP will identify such a person and will NOT apply the test to them.  I believe that DNAP has demonstrated that mouse DNA will not work in the test.  Would chimp DNA?  I do not know.  If there are sufficient differences to human, there will be too many “failed” markers, and the test would not work.  DNAP probably assumes that those taking the tests are Hominids and not Pongids.

6. The Hawksian addendum is sort of irrelevant, as DNAP looks only at population distinctive markers with a large “delta value.”  They do not look at those with frequencies so close as to be meaningless.


Posted by JW Holliday on Mon, 08 Aug 2005 17:39 | #

For the critics

With the following assumptions that I view as absolutely fundamental to the test:-

1) Looking only at statistically significant results (keeping in mind sample size as well)

2) Understanding that what is being looked at it relative genetic affiliation, without the (absolute) requirement of direct ancestry.

3) That labels of genetic affilations do not always represent the totality of the groups themselves.

4) The test is unable to fully distinguish Native American and East Asian when those ancestries are a relatively minor component.

With those 4 assumptions, what, pray tell, are the large and obvious errors inherent in the DNAP data as it currently exists for various ethnicities?

Certainly, if the test itself - rather than certain naive interpetations - is “wrong” and “lousy”, then certainly one could point to obvious problems with the data, things that simply do not make any sense, things that indicate random or systemic error rather than a correlation of data with known population histories.

Please, without a priori expectations and WITH acknowledgement of the 4 fundamental points listed above, if the tests are wrong, where is the evidence?

Why are North Africans predominantly Caucasian with a strong African minority component?  Why do the data for African Americans and Puerto Ricans match well known population history and independent data in the literature?  Why do the Northern and Southern European affilations match well Cavalli-Sforza’s component 1 gradient of European genetic variation? Why the gradients of Middle Eastern ancestry?

In other words, putting aside statistically insignificant results and the confusions between labels and reality, where’s the errors?  Where are the clearly identifiable mistakes, not only in majority affilation, but in the statistically signficant minority affiliations as well?

“Where’s the beef?”, with respect to these gross errors?

Support the contention that the tests are wrong and lousy based upon the empirical evidence of the results.


Posted by JW Holliday on Mon, 08 Aug 2005 17:42 | #

I have discovered that Dienekes is incorrect about the chimp comment.  From we

Number of failed Loci

Chimpanzee - 157 out of 176 Gorilla - 151 out of 176 Orangutan - 137 out of 176”

Therefore, a chimp would NOT give a ABD2.5 result, the overwhelming majority of loci do not work.  And, while African DNA would “work” with Euro 1.0, the algorithm would screen it out.

This chimp stuff is a minor point, but the underlying implication is important.  Which is, many of the criticisms of the DNAP test are due to people either not understanding the tests and/or not doing carefully looking at the online material that demonstrate that some criticisms are simply incorrect.

I wonder as well….when more advanced tests are developed (by whomever), and when these tests confirm that the current DNAP data is essentially correct, will the critics admit their error?

Certainly, if the new data refutes the current DNAP data, I will admit that fact, and openly revise my opinions as a consequence.


Posted by manju on Tue, 09 Aug 2005 05:39 | #

Looking at populations of northern and eastern India, Nepal, Bangladesh, Burma etc, I fail to see why it is so impossible that South Asians are a Caucasian/Mongoloid mix,
If you mean North-East India then you are judging 97% of the population in India based on 3% Indians. Also, majority of North-Eastern Indians have recent migration history to that area from South-East Asia.

Regarding East Asian admixture in South Asian population, I would agree with Dienekes words;
They are descended from very old indigenous South Asian populations as well as more recent Central Asian (Caucasoid) populations. Whatever similarity they have with East Asians is due to common ancestry _before_ the emergence of Mongoloids.

However, Central Asia also incldues North-West India, Pakistan and Southern Afghanistan which are part of South Asia. Considering the following three factors,
1. Old indegenous population
2. Common ancestry with East Asians
3. Part of a region which is regarded as geographical boundary that gave rise to Caucasian features,
            South Asia instead of being subset of Caucasians must be the superset of East Asian, Central Asian and Indian. The term Caucasian should be made obsolete. However, better term would be ‘Indian’ as a supergroup of which East-Asian, Central Asian and South Asian are the subgroups.
                    Instead of this Indo-centric view, we should have two dispersal regions.
1. The primary dispersal region: Africa
2. The secondary dispersal region: Central Asia and the Indian subcontinenet.
Then the movement of Homo Sapiens would be,
Primary dispersal region -> Secondary dispersal region -> Other parts of the world.


Posted by JW Holliday on Tue, 09 Aug 2005 11:47 | #


I advise you to look at the links regarding population genetics of South Asian populations.

Dienekes, regarding his update. First, Dienekes, as is typical of DNAP critics, cannot separate the names given to categories from the categories themselves.  Just because DNAP print calls Caucasian “European” does NOT mean that “South Asian” is a West Eurasian ancestry.  There is no reason that it does not represent the aboriginal South Asian ancestry.

I wonder if Dienekes has ever considered that some of the “bugs” in the test are simply due to inadequate sample size.  If someone wants to say - 17 Irish are insufficient, I’ll agree.  But until 100 Irish are compared to 100 Turks, I’m withholding judgement.  And, IF someone wants to “make hay” over current results, why only pick out the negatives?

You see, this is the REAL problem I have with the critics - they view it as black/white and not gray.  The tests are not perfect, which is why DNAP is updating them.  But to dismiss all the data as wrong/meaningles/bugs based on cherry picking some (often statistically insignificant) numbers while ignoring the broad range of data, indicates some underlying animus.

Forgetting about chimps (a quite inane example), and moving to the more realistic example of Australians.
Actually I agree with Dienekes here.  Of course, problems with an isolated race does not logically imply that the major racial groupings suffer from the same problem.

However, I agree that DNAP needs to extend its parental populations.  I in fact have never been quite pleased with that aspect of the test.  That, however, does not invalidate the bulk of the data, in a comparative sense.

It should NOT be called an admixture test, or even a direct ancestry test, it is a test of relative genetic affiliation, based upon those categories measured.  There is room for improvement.

As regards the different possibilities of Mongolid affiliation in South Asians, Dienekes should read the “intepretations” section of the ABD site, since what he is saying has essentially been admitted by DNAP long ago - although DNAP typically do a poor job in explaining themselves.

I will congratulate Dienekes ... his points, particularly about the parental populations (which I agree with, BTW), are the ONLY critiques of DNAP that have any value.  Although why these critiques started only after DNAP started talking about Greeks ...

It is of interest that a non-specialist like Pontikos at least makes his points, while Hawks (and a certain other person) sounds like he doesn’t know what he talking about. 

I take a glass half full approach.  I think the DNAP tests give much useful information. The emprirical data, in toto, support the validity of the tests, although refinement is needed and DNAP needs to be more forthcoming.  However, the issue of Australians and small sample sizes aside, the tests work.  IF properly interpreted.

My question to the critics - if your criticisms are based on science and not politics, have you addressed these concerns directly to DNAP?  If yes, what was their answer?  If not, why not? Particularly to the ‘brilliant’ academic ‘experts’ who critique the test (and sound like asses) - why not offer your brilliant opinions and (paid) services to DNAP as consultants?

And when the hell is DNAP going to start defending their own product?  Publish papers on the full methodology, release more data, make the photodatabase open to the public….my keeping silent, DNAP is making themselves look bad.


Posted by JW Holliday on Tue, 09 Aug 2005 12:00 | #

A few further clarifications, particularly in response to Dienekes Pontikos.

1. See here particularly mechanism 3 with respect to South Asians.

As DNAP has *already admitted* that certain populations may be intermediate to those used to calibrate, and have *already admitted* that South Asians may be such an intermediate population, “novel” discoveries of this as a critique do not impress.  Furthermore, there are population genetic studies that claim Mongolid influence in South Asian populations, including North Indians.  Now, these may be wrong (I myself am skeptical of HLA studies), but one cannot say with certainty that South Asians do not have Mongolid influences.  Do you really know that for sure?

2. If DNAP wishes to defend their product as an admixture test, then *they* must do that.  I will not.

I will defend the test as a measure of relative genetic affiliation to the parental populations.  Note that given this interpretation, one can take DNA data from Australians (or chimps) and determine relative genetic similarity toward combinations of the parentals.  Granted, for Australians and chimps this would not be very informative.  However, I myself do see it as informative, for most of the world’s human populations to determine affinity toward the 7 total categories.  I myself find much useful information there.  If others do not, that is their preference; but the test does work as a measure of this affiliation, and there are people interested in such comparative studies. And there are phenotypic correlations, as well as some historical ones, with these data.

3. Besides too small sample sizes, the other problem in using DNAPrint’s ethnic data to bash the test (or admittedly, to overly praise the test) is that many of the categories are vague.  Turks.  From where?  Western Turks and those from Ankara will most likely be quite different than those from the East or Cappadocia.  Northern Europeans - from where?  South Asians - from where?  Iberians?  Portugal?  Spain?  Where in Spain?  Catalonia?  Andalusia?  Italy - North, Central, South, Islands?  Ireland - East or West?  What are the two Middle Eastern populations?  And so forth…

Let there be no mistake.  While I strongly support the DNAP tests as measures of relative genetic distance/affiliation, I am not at all pleased with how DNAP itself is handling the tests.  So much unexplained, so little concrete information (including about the parentals), so poor explanations.  They need to do a better job to justify their “low, low prices”, I agree on that.  But this does not logically mean that all the data from the test are meaningless.  The critics are fundamentally flawed (and exposed as somewhat irrational) because they refuse to take a nuanced view (as do more reasonable people like Steve Sailer and Nusapiens) and balance the positive and the negative.  To the critics, it is all negative, and that’s not the case.

4. I’ve repeated endlessly about the categories and what they mean (and even Hawks seems to agree on that), so I hope the critics will incorporate that into their thinking.


Posted by JW Holliday on Tue, 09 Aug 2005 12:03 | #

With respect to the issue of relative genetic affiliation (not admixture per se), “mechanism 3”, and South Asians, I’m wondering why no anti-DNAPer has of yet negatively critiqued the Ray et al paper “Inference of human geographic using Alu indertion polymorphisms”, in Forensic Science International.

This paper infers probabilities of ancestries (using Alu sequences, not SNPs)from one of 4 base populations - Africa, Asia, Europe, and India (note India is so different from Europe it must have its own category).

One reads such comments as: “Sample 15 tested as an admixture of 86% European/10% Asian, which is consistent with a Venezuelan origin.”

Note that they are inferring *admixture* from their data (the horror!!!), and as well are assuming that New World Amerind ancestry is in the “Asian” category.

Note that that their South Asian Indian subject had a 49.1% probability of a South Asian classification, and a 45.9% for Asian (here, “Asian” means East Asian, a Chinese was 93.8% probability Asian).  Interesting.  Thus, even with a separate South Asian “Indian” category, a trend toward “East Asianess” was apparent.

As well, the authors state:-

“...India is by far the most heterogeneous with many individuals clustering with either Europe or East Asia (referencing a Bamshad paper).”

And, while the authors, in the abstract note homoplasy (identical by state, not descent) as a possible problem with a SNP-based method, in the text they note:-

“Recent improvements in SNP-based approaches suggest that most of these problems can be overcome by carefully selecting the correct loci and the correct number of SNPs to use.”

The authors also note they are trying to identify markers to give sub-continental classifications.

This seems to have the same advantages and disadvantages of the DNAP methods, albeit less developed at this point.  So far, the two seem to give roughly similar results.

Yet, DNAP is attacked, and Ray et al is not.  A clue to DNAP: PUBLISH your methodology and data.  As well, the fact that DNAP is commercializing this may offend some.

One wonders what will be done with the Alu method, as one of the funders of the work was the CIA.


Posted by Jens on Tue, 25 Apr 2006 08:43 | #

Does anyone know the the genetic distance between east asians (japanese, korean, chinese) and chimps? I have heard that Karachi mongoloids are pretty close to chimps in terms of genetics. At a site called EZ-board the claim is that nigerian genes closely resembel that of chimpanzees. Is there a bigger genetic difference between europeans and africans(eg. sokoto nigerians) than there is between east asians (japanese(koreans/chinese) and africans (eg. sokoto nigerians)? I have heard that in some old studies East asian are furthest away in terms of genetic distance from africans, but i have also heard that more recently studies (more comprehensive studies) puts the east asian like the japanese and koreans closer to the africans. 



Posted by ben tillman on Tue, 25 Apr 2006 10:34 | #

Is there a bigger genetic difference between europeans and africans(eg. sokoto nigerians) than there is between east asians (japanese(koreans/chinese) and africans (eg. sokoto nigerians)?

I’ve read that the genetic distance between a German and a chimp is only four times as great as that between a German and a Nigerian, but I don’t know about East Asians.


Posted by Jens on Tue, 25 Apr 2006 11:44 | #

Do anyone know, maybe Ben, where I can find some updated data regarding genetic distance between various human groups and perhaps some comparison chart with chimpanzees? - Data which are comprehensive (involving a substantial number of locus) to give a fair comparison and real picture of the relationship between different human groups and evolution. 


Post a comment:

Name: (required)

Email: (required but not displayed)

URL: (optional)

Note: You should copy your comment to the clipboard or paste it somewhere before submitting it, so that it will not be lost if the session times out.

Remember me

Next entry: Affirmative Action Lunch
Previous entry: The Bear’s Lair: The jobs aren’t coming back

image of the day

Existential Issues

DNA Nations

Establishment Problem



Endorsement not implied.


Islamist Threat

Anti-white Media Networks






Historical Re-Evaluation

Controlled Opposition

Nationalist Political Parties


Europeans in Africa

Of Note


Benefit commented in entry 'These adverts communicate an important message, telling us the aims and desires of the corporations' on Mon, 20 Nov 2017 05:10. (View)

mortgages commented in entry 'These adverts communicate an important message, telling us the aims and desires of the corporations' on Mon, 20 Nov 2017 04:12. (View)

Whitney Sara Cohen commented in entry 'These adverts communicate an important message, telling us the aims and desires of the corporations' on Mon, 20 Nov 2017 03:33. (View)

Jean Hill commented in entry 'JFK assassination files released: declassified documents reveal CIA plots to kill Fidel Castro' on Sun, 19 Nov 2017 21:41. (View)

Captainchaos commented in entry 'These adverts communicate an important message, telling us the aims and desires of the corporations' on Sun, 19 Nov 2017 13:27. (View)

Origins of fake pizza news commented in entry 'When the Alt-Right needs to subvert a fledgling White Left, they turn to Nowicki, Preston and Stark' on Sun, 19 Nov 2017 10:56. (View)

Prince, Pettibone, Posobiec commented in entry 'When the Alt-Right needs to subvert a fledgling White Left, they turn to Nowicki, Preston and Stark' on Sun, 19 Nov 2017 10:53. (View)

Origins of "Pizzagate" fake news: interesting commented in entry 'When the Alt-Right needs to subvert a fledgling White Left, they turn to Nowicki, Preston and Stark' on Sun, 19 Nov 2017 10:48. (View)

DanielS commented in entry 'These adverts communicate an important message, telling us the aims and desires of the corporations' on Sun, 19 Nov 2017 08:23. (View)

Guessedworker commented in entry 'These adverts communicate an important message, telling us the aims and desires of the corporations' on Sun, 19 Nov 2017 06:04. (View)

Backgrounding/Jamming race-mixing agenda commented in entry 'WHITE WOMEN FOR SALE!' on Sun, 19 Nov 2017 01:00. (View)

William of FN commented in entry 'French Court Rules: No Such Thing As Indigenous French' on Sat, 18 Nov 2017 13:45. (View)

Trump Jr. > Wikileaks > Kremlin commented in entry 'There's no "there there" to Russian investigation only where the Alt-Right doesn't want there to be.' on Fri, 17 Nov 2017 22:17. (View)

Elizabeth I commented in entry 'PM speech to Lord Mayor's Banquet stresses importance of international rules-based system' on Fri, 17 Nov 2017 08:15. (View)

Bush, Ford and Warren commented in entry 'JFK assassination files released: declassified documents reveal CIA plots to kill Fidel Castro' on Thu, 16 Nov 2017 20:56. (View)

Bush / JFK assassination connections commented in entry 'JFK assassination files released: declassified documents reveal CIA plots to kill Fidel Castro' on Thu, 16 Nov 2017 20:25. (View)

More earthlike planets being discovered commented in entry 'NASA invests in 22 visionary exploration concepts, including asteroid mining' on Thu, 16 Nov 2017 12:55. (View)

Kushner co. accused of illegal rent inflation commented in entry 'Undoing inherited wisdom & means of separatism / forcing integration - YKW doing as YKW do' on Thu, 16 Nov 2017 04:29. (View)

Louis CK and the 1/4 issue commented in entry 'Test Your Capacity To See Through Jewish Crypsis: Which ones are Jewish?' on Thu, 16 Nov 2017 02:49. (View)

Un-Fairstein comes to Weinstein's rescue commented in entry 'Coerced Confessions of The Central Park Five' on Thu, 16 Nov 2017 01:43. (View)

Angela Nagle discussing her take on the Alt-Right commented in entry 'Angela Nagle: they think voluntary outbreeding is genocide. Cultural Marxism, Jewish porn! lol.' on Wed, 15 Nov 2017 13:09. (View)

African knowingly gives HIV to French women commented in entry 'black hyper-assertiveness' on Wed, 15 Nov 2017 11:04. (View)

Spencer's straw man commented in entry 'Poland is correct to denounce Richard Spencer' on Wed, 15 Nov 2017 10:42. (View)

Black murder of White working class commented in entry 'black hyper-assertiveness' on Wed, 15 Nov 2017 01:37. (View)

Evidence of Trump-Russia collusion exists commented in entry 'Abnormal: celebrates the Soviet Union's victory and fêtes Russia's imperialist legacy.' on Tue, 14 Nov 2017 10:50. (View)

Turkey/Flynn plot to kidnap Gulen commented in entry 'Mueller investigating Flynn over alleged plan to deliver accused coup organizer to Erdogan' on Tue, 14 Nov 2017 10:41. (View)

Key moment in decline of US power in Asia-Pacific commented in entry 'North Atlantic: You Have Spread Your Dreams Under Their Feet' on Tue, 14 Nov 2017 10:13. (View)

Not true, Putin told me so. commented in entry 'Abnormal: celebrates the Soviet Union's victory and fêtes Russia's imperialist legacy.' on Tue, 14 Nov 2017 06:59. (View)

DanielS commented in entry 'Pragmatism as ethnonationalism's tool against radical skepticism' on Mon, 13 Nov 2017 22:09. (View)

Arthur commented in entry 'So what are we to make of the crisis in Catalonia?' on Mon, 13 Nov 2017 20:12. (View)

Guessedworker commented in entry 'Pragmatism as ethnonationalism's tool against radical skepticism' on Mon, 13 Nov 2017 18:00. (View)

Guessedworker commented in entry 'Pragmatism as ethnonationalism's tool against radical skepticism' on Mon, 13 Nov 2017 15:47. (View)

Palace intrigue in Saudi Arabia and Lebanon commented in entry 'Trump may have pushed Saudi Arabia and Iran closer to war' on Mon, 13 Nov 2017 13:31. (View)

DanielS commented in entry 'Pragmatism as ethnonationalism's tool against radical skepticism' on Mon, 13 Nov 2017 11:41. (View)

mancinblack commented in entry 'U.K. Cabinet Minister Priti Patel Resigns Over Secret Israeli Meetings' on Sun, 12 Nov 2017 09:06. (View)