The Study of Racialism

I am wondering if the cDe blood protein (or antigen) would be considered a wimpy marker of African ancestry, or a strong one.

One old 1962 source (a book) someone posted, Dzobhansky (I think), claims that since it is so widespread, one cannot assume that everyone who possesses it has a Negro ancestor just because it is found in its greatest volume in sub-Saharan Africa. It seems to me that Dzobhamsky misses the point. If it reaches its greatest quantity in sub-Saharan Africa, the logical deduction is that it originated there.

The study we were discussing the other day from 1978 (Sandler, et al.) definitely considers it a sub-Saharan marker, as do a few others I've read, including one very recent one that says it is predominantly African...which definitely suggests a sub-Saharan origin. You also said that all the markers mentioned in the 1978 study are definitely sub-Saharan in origin, so I'm assuming that includes cDe.

It seems to me, and correct me if I'm wrong, that it did indeed originate in sub-Saharan Africa and thus is an indicator of sub-Saharan ancestry, despite its widespread presence, even in Eskimos (according to something I read). It may just be very old and have spread a long time ago. Its age would play havoc with trying to use it to quantify modern-day admixture in populations, but its age can't somehow magically render it as non-sub-Saharan. But would its age and the fact that it is widespread make it a wimpy, albeit valid, marker?

That is a damned good question, and I wish I could give you an equally good answer.

Let me first explain, for the rest of our readers, that everyone's chromosome #1 has three adjacent genes called C, D, and E. Each gene comes in two flavors (alleles) traditionally denoted as upper-case and lower-case. And so, each chromosome #1 has one of eight different combinations (haplotypes) of upper- and lower-case genes: CDE, CDe, CdE, Cde, cDE, cDe, cdE, or cde. You have two chromosome #1's, one from your mother and one from your father, and they can be of different haplotypes. The three genes are medically important because a foetus with two "cde" chromosomes (this is called "RH-negative") is likely to die in utero (perhaps killing the mother as well) if she has a chromosome with the upper-case D (this is called "RH-positive").

Here is the history, as I know it [taken from Luigi Luca Cavalli-Sforza, Paolo Menozzi, and Alberto Piazza, The History and Geography of Human Genes, trans. Sarah Thorne (Princeton: Princeton University, 1994) pages 132-133.]

1. As far as anyone can tell, our ancestors before the great African diaspora of 60 kya all had "cDe." This was the original haplotype. Hence, it is more common in sub-Saharan Africa than anywhere else.

2. Then, a "C" mutation in the band that left Africa changed some of them to "CDe" but left others with "cDe." It is found mainly in Asia, and in about 9 percent of Africans.

3. Next, a "D" mutation among the the migrants who were still "cDe" changed some of them to "cde." These were the guys who populated Europe. This haplotype ("cde") is not rare nowadays in sub-Saharan Africa, however. This means that it may have flowed back to Africa with the Phoenecians, Egyptians, Romans, etc. Incidentally, "cde" is rare in Europe today because most of the European hunter-gatherers were replaced by farmers from the middle east about 10 kya. The Basques are the sole remnant of the original European population.

4. Finally, a "E" mutation among the Asian migrants who were still "cDe" changed some of them to cDE. These migrated to northeastern Asia and then to the Americas ("cDE" is also found in southeast Asia and Australia).

5. The other combinations seem to have arisen by crossovers between the above types.

Your question comes down to two separate issues. First, does the high incidence of "cDe" in Africa mean that it came from Africa? I would say yes, definitely. Second, are the "cDe"s found around the world today simply remnants of the original African diaspora or do they reflect recent gene flow (within the past 10ky)? I would say that they do not reflect recent gene flow, but are remants of the original diaspora. I say this because "cDe" is so widespread. But I honestly do not know how you could prove it either way.

That is an excellent answer, as it gives me loads of information I didn't have before. Thanks.

Incidentally, is cDe(Rho) the same as cDe?

Yes. Karl Landsteiner and Alexander Wiener discovered the Rh blood types in 1940. They did not know the underlying "CDE" genetics of the system, but they adopted the term "Rho" for what we now know is any haplotype with an upper-case "D" (the most common being "cDe").

So, being that cDe is the marker of all our ancestors before the diaspora, it is sub-Saharan in that sense. Thus, its presence in non-African populations is most likely due to the original diaspora, even though some of it could have been introduced by later migrants. There is no real way to tell. Gotcha...

In that Sicily study from 1978, where HbS, Fy(a-b-), cDe(Rho), V(hrv) and Jsa were found, do we forget about cDe, or does it perhaps have some significance here because of the other more recent sub-Saharan markers mentioned above along with it (and because L lineages were found by other researchers)?

Separate question: are all L lineages sub-Saharan in origin? I'm aware of L1, L2, and L3. Are these all sub-Saharan? Are there more besides those, like L0 or L4? And how about M1? I've seen this described as sub-Saharan by some, and not so by others.

I don't know. You could argue it both ways. Personally, I give a lot more weight to the "L" mtDNA.


Yes, all of the Ls are sub-Saharan in origin. And yes, there are more, but not exactly in the sense you said. There was a lot of Eurocentrism in the naming system. They divided Europe and western Asia into nine categories H, I, J, K, T, U, V, W and X. They divided eastern asia (including native Americans) into eight categories A, B, C, D, E, F, G and M. But they assigned only one category (L) to all of sub-Saharan Africa.

Now, given that there is far more diversity in sub-Saharan Africa than in the rest of the world combined, this makes as much sense as dividing the animal kingdom into thirteen categories: a dozen labels for different breeds of dog and one label for all other creatures on earth.

And so, as people began to study Africa (funding there is hard to get), they subdivided L into the three main groups you mention (L1, L2, L3). Given the nature of taxonomic naming schemes, I would not anticipate the discovery of L0 or L4, but I would predict further subdivisions into L1a, L1b, L1c, etc., where each subdivision is as important and richly varied as, say, H or J.


That is one of the burning questions of molecular anthropology today. I recall a class of geneticists and anthropologists debating this, hammer and tongs, for about a week. M1 is found on the eastern edge of Africa (Somalia) and all over southern Asia. Theory 1: It appeared in Somalia 60 kya, inside a small sub-group within the band who then launched themselves across the Straits of Bab el Mandeb to colonize the planet. Theory 2: It appeared much later in southern Asia and then leaked back to Africa. Evidence for theory 1 is that all the other Ms seem to have sprung from M1 by mutation, so it must have been first. Evidence for theory 2 is that it has highest incidence in southern Asia, so southern Asia must be its original source. I refuse to take sides.

Thanks, Frank, for all this information is definitely giving me a clearer picture of everything.

1) One last question about that Sicilian blood group study: We know that Hbs (sickle-cell) and Fy(a-b-) (duffy null) are of relatively recent sub-Saharan origin, and thus indicate recent sub-Saharan gene flow, right? We also know cDe(Rho) is ancient, and its presence in Sicily may or may not be an indicator of recent sub-Saharan gene flow. How about the antigens V(hrv) and Jsa? Judging from the study itself, it looks as though they must be of relatively recent sub-Saharan origin, since the researchers used them. Do you know anything about these in particular?

2) Regarding the L mtDNA haplogroups, what does it mean when they, like L3, for instance, are given lower-case letters, such as L3e and L3a? How about when asterisks follow them, as in L3*? I agree with you that it seems stupid to assign the population with the most diversity a single letter and then have to sub-divide it.

3) Interesting what you wrote about M1. I guess it wouldn't be wise to use it as an indicator of relatively recent sub-Saharan ancestry, but it looks like it possibly may be able to be used to indicate non-European ancestry.

4) What do you think Sykes used when he studied the British population and stated that one out of every 100 Britons has an African or Asian ancestor? Here is a link (I can't find much other info, though, other than reports such as this):

http://www.skfriends.com/black-dna-in-white-british.htm

[edited]

Here is an article with more info:

http://www.freedomsite.org/pipermail/fs ... 01866.html

5) Lastly, what do you think of HLA studies, like the one in the link below? I recall when it was first posted on a message board, white-nationalist Greeks were up in arms, and tried to discredit it in every way (unsuccessfully), even resorting to claiming it had been retracted, when it indeed had not. Another study by the same reseacher had been retracted, but it concerned relatedness of Palestinians and Israelis, even though some of the Greek-Ethiopian data was cross-referenced. The study of Palestinians and Israelis, by the way, in my humble opinion, was retracted for political reasons, regardless of what they claim. Anyway, here is the Greek-Ethiopian study:

http://www.ncbi.nlm.nih.gov/entrez/quer ... query_hl=1

Sorry. As I recall, the V antigen is designed to kill certain types of rotaviruses. I have no idea what the Jsa antigen does. Since most blood protein antigens (blood groups) vary among populations, they used to be very popular as markers before the decoding of the genome. They have become less popular with the identification of mtDNA makers and, later, of the many SR-Y and autosomal markers.


L3a, L3b, etc. are simply breakdowns of L3 into subgroups. I am not specifically familiar with the designation “L3*” but if you found it in an article about migrations, it probably refers to the ancestral L3 before the sub-groups (L3a, L3b, etc.) mutated and split off. L3 is important because it is the ancestor of M (and all other non-African mtDNA haplogroups), meaning that the original out-of-Africa band had L3 mtDNA (and perhaps some M1 also). Regarding the taxonomy: hindsight is great, but when it was devised, nobody had any idea that Africa had so much diversity, nor that it was our species' original homeland, for that matter.


I agree. By the way, if you are interested in mtDNA (which has a great track record since it has been around longer than other molecular methods) I have heard good things about The Real Eve: Modern Man's Journey Out of Africa by Stephen Oppenheimer. Also, here are some mtDNA phylograms and maps that I had lying around.







I think what he is saying is that one percent of Brits have an African or Asian mtDNA haplotype. (Sykes specializes in mtDNA.) But this is one of those meaningless attention-getting hooks that article-writers use. I find it hard to believe that Sykes was quoted correctly, but it is meaningless anyway. You had a million ancestors in Columbus’s time, and up to a billion ancestors in William the Conqueror’s time (more than the planet’s population then). The statement says that one of those ancestors had African or Asian ancestry. Give me a break!


Well, I have never really gotten into HLA stuff. It is hard to say why, since I started to look into it a few years ago, and many people whom I respect swear by it. I guess three things turn me off about it. First, the nomenclature seems needlessly complicated somehow. Second, nobody has come up with a molecular clock for the rate of HLA mutations. This is because HLA is the section of the genome (on chromosome #5) that is specifically designed to produce antibodies. And so, it mutates at an astronomical rate (organisms that cannot keep up with the germs in the eternal arms race go extinct). Worse, different sections of the MHC mutate at different rates. Finally, certain HLA lineages appear in the other hominids. In other words, certain modern-day human HLA lineages separated from each other thousands of years before we separated from chimps. I have a hard time getting my mind around that last finding.

Thanks for the charts and diagrams...I'll study them when I have a little time.



Does this mean that when L3 is found in Europeans or other non-Africans, it isn't necessarily indicative of sub-Saharan ancestry?

You mean, might it be left over from the original diaspora? I do not think so. It is not found in any concentration anywhere outside Africa. Every haplotype that descends from L3 (but not L3 itself) shows definite gradients (called "clines") outside of Africa, indicating where it first popped up. The thinking is that the original band mutated to M1, and then gradually to all the other non-African haplogroups, and that no ancestral "L" maternal lineages survived outside of Africa.

Thanks for the L3 explanation. In the diagram you provided, I noticed L1*, L2*, and L3* are in the center and the others branch off of it. Like you said, that must mean they are the "parents" of the others. How come in one of the charts an "L0" is listed, all the way at the bottom? Is this a hypothetical one?



So, generally, counting up the number of African mtDNA haplotypes (or Y-chromosomes, for that matter), in Europeans isn't a good way to estimate black admixture percentages, as you explained with the sock-drawer analogy. They simply prove sub-Saharan DNA is present, thereby proving that there were black ancestors, however distant. People on other message boards used to tally Europeans' black admixture by doing this, but I never saw any wisdom in it. So, to say that Sicilians have 4.4% black lineages (in one mtDNA study), and Portuguese have 7% (in another mtDNA study) is not to say that these respective populations are 4.4% or 7% black, or that they have that amount of black admixture, correct?



But this doesn't suggest they are now suddenly considered totally useless for tracking population movements or the existence of admixture, correct? I think one "white nationalist" type of person on a message board was trying to suggest this.

As I recall this is a chart by Dan Mishmar of U of Cal, Irvine, who proposed labeling the haplotypes that most people call "L1a" and "L1b" as "L0" and "L1" instead. I do not recall why he wanted to do this.


Correct.


It depends. How much confidence do you have in the randomness of your sample? All of the following explanation assumes that the people whose DNA was tested were chosen randomly from the general population.

Go back to the drawerful of socks analogy, but now with an entire population of drawers (millions of drawers, each with two hundred socks). You are going to pull only one sock (the mtDNA sock) from several of the drawers. From how many drawers should you pull the mtDNA sock, in order to say that those few drawers are truly representative of the millions of other drawers? Ten? A thousand? A hundred thousand?

It depends on how sure you want to be. If you look at the mtDNA socks of 1,000 drawers and you count 7% black socks, the chances are 95/100 that the entire population will have, on average, between 5.9% and 8.1% black socks. But if you look at the mtDNA socks in only 100 drawers, then the 95 percent confidence interval expands, and you can say only that the entire population will be somewhere between 3.5% to 10.5% black socks. The bigger the sample, the tighter the confidence interval will be.

The percentage that you find also affects the confidence interval. Say that you examined the mtDNA sock in only 100 drawers and found 50-50 white/black. You would then be 95/100 confident that the total population was actually between 43% and 57%. But if you examined the mtDNA sock in 100 drawers and found no black socks at all, then you could be 95/100 confident that the whole population (every sock, every drawer) was entirely white.

Let's go back to the Portuguese case. If the 7% "L" haploytpes were in a sample of 1000 individuals, it shows that between 5.8% and 8.1% of Portuguese DNA is African. But if the number came from sampling only 100 individuals, then it shows that between 3.5% and 10.5% of Portuguese DNA is African. Finally, if only fifteen individuals were sampled, then it shows that between zero and 15% of Portuguese DNA is African (meaningless).

As it turns out, I personally rely on such studies to show the extent of African admixture in Iberia (Portugal and Spain) because they match historical data documenting the importation of about 100,000 African slaves in the 1500s, who, according to first-hand accounts, were then assimilated and absorbed into the Iberian population.

I have read statistics that during the course of slavery in Portugal, beginning in 1441, about 35,000 black Africans lived there, with many of them having been absorbed. I haven't read anything specifically on Spain. I guess stats will vary.

What baffles me is when I see written in historical accounts that black Africans are one of the major components of the Portuguese people, when that makes no sense at all, when considering the importation figures and mtDNA, and when considering the physical attributes of the population...i.e., they don't have Negroid phenotypes at all.

What baffles me even more is when I see written that Portuguese have Negroid physical features...an outright lie, or a mistake made by someone who obviously hadn't travelled extensively there. Perhaps those who wrote this visited Lisbon or Evora in the 1600's and saw some mulatto types (since both of these cities had a 10% black population in 1600 or so), and made this assertion, not realizing how few mulattos / blacks there were in relation to the white population on the whole, thereby not realizing that this small African population couldn't make the Portuguese population visibly biracial. Anyway, saying the Portuguese are visibly mixed is nonsense.

I've read that during the 1600s and 1700s, London and several other British cities had about 20,000 Negroes each. I've also read that many Blacks existed in Holland at this time, and in France, and throughout peninsular Italy. I wonder why blacks aren't considered a "primary component" (how ridiculous) by those same writers in these populations.



So, what would the Sicilian confidence level be? The mtDNA study I'm referring to is the one that found 4.4% black African sequences (HpaI-3 [equivalent to the +3592HpaI] AvaII-3 association, corresponding to haplogroups L1/L2). They found four such people out of 91 studied. I'm reasonably certain that the mtDNA percentage would equal the percentage of blacks in Sicily during the time of the Saracens, since I've read that blacks were not uncommon in Sicily at that time.

Assuming that the 91 individuals tested were a random sample of a population, a 4.4% rate in this sample gives a 95/100 confidence interval that the actual African genetic admixture in the entire population is between 1.4% and 7.4%.

The 95 percent confidence interval formula is: PE=1.96*SQRT(r*(1-r)/(2*n)) where:
PE is the possible error.
r is the measured rate.
n is the sample size.

The confidence interval ranges from a low of r-PE to a high of r+PE.

Hi Frank,

I was seeing the picture of the spread of human populations throughout the world.

As you know, I am very interested in the origin of the American Indians, and once I said in another thread that the Siberians were their closest people living in the Old World. I was surprised at that time you say the theory of the relation between Siberians and American Indians was discarted by Science some time ago.

How come?

As far as I can see in the maps you posted the routes of the spread of humans from Eurasia to America cross Siberia, so the peoples of those regions should be (by intuition only) related to Native Americans.

If not Siberians what other peoples living peoples of the Old World are the closest ones to Native Americans?

I am very curious, indeed.

Regards,

Omar Vega

Today’s Mongolians are the closest relatives of the group that crossed from Siberia to the New World. Let me tell you the story.

Connie Kolman (now Connie Mulligan since she got married a few years ago) is one of my molecular anthropology professors at the University of Florida. This topic came up in class and, as it turns out, she was the very person who had solved the puzzle. To understand the solution, we need to define the puzzle.

The problem is not from where the Native Americans came. There is no doubt that these mammoth hunters crossed Beringia from Siberia about 20 kya when sea level was lower. This is not to say that other individuals from the Old World did not immigrate in boats. Merchants do tend to wander about. For instance, after humankind’s 100 ky of hunting-gathering, people at four widely separated spots around the globe (the Middle East, China, West Africa, and Central America) simultaneously invented agriculture. Simultaneously! Clearly, knowledge of the new technology flashed around the planet in just a few centuries. The only thing you need to know to re-invent most technologies is the knowledge that someone else has done it.

The puzzle is which group of today’s Asians are the ancestors of (more precisely, share the most recent common ancestry with) today’s Native Americans. Native Americans have a very specific mix of A, B, C, and D haplotypes. Each of these four haplotypes is found also in Asia (and nowhere else), but they are widely separated. A, C, and D are found in regions with no B. Populations with the B haplotype are found far from those with A. Worse yet, in Asia, each of the four Native American haplotypes is mixed in with G, M, Y, and Z haplotypes (which are unknown in the Americas). Until Connie’s project, the only explanation was that A, B, C, and D individuals somehow separated themselves from their G, M, Y, and Z neighbors, friends, and relatives and then trekked across the globe (perhaps in separate waves).

Connie (photo above) spent two years collecting DNA from the many tribes throughout Mongolia. She discovered that Mongolians not only have the same mix of A, B, C, and D, but that they lack the other common Asian haplotypes. Her explanation (now widely accepted) is that Mongolian mammoth hunters followed their prey across Beringia during the last glaciation. Later, when the climate warmed and the mammoths were exterminated, their relatives in Asia switched to nomadic pastoralism on horseback. Meanwhile, Siberia was repopulated by unrelated migrants from northwestern Asia, descendants of the Saami of Finland.

When she told the story in class, I was amazed at the fortitude and courage of the delicate-looking lady. Just imagine her living in tents and drinking fermented mares' milk for two years! With a grin, she explained that she had done nothing of the sort. She had worked for two years in an office at the national university in Ulan Bator (Ulaanbaatar), which enrolls students from all over Mongolia. She simply collected DNA from the students, recording each one’s tribe and home province.

Dr. Mulligan's project was published as: Connie J. Kolman, Nyamkhishig Sambuughin, and Eldredge Bermingham, “Mitochondrial DNA Analysis of Mongolian Populations and Implications”, Genetics 142 1321-1334 (April, 1996). You can download a copy from http://backintyme.com/admixture/kolman01.pdf.

I know you don't have much faith in the reports of the Sykes study on Britons, but if he indeed did test for mtDNA in 10,000 Britons (excellent sample size), and found one percent of them to be African or Asian, doesn't that mean we can plug in those figures into the 95% confidence interval formula and get some sort of useful data on African or Asian admixture with a relatively narrow range?

The 95% confidence interval forumla, by its very name, of course, doesn't purport to give definite admixture figures or ranges, only what we can probably expect, correct?

I've seen another method employed on message boards, although never by geneticists, on determining overall admixture from mtDNA. These individuals in question take the percentage of Negroid mtDNA found in a European, and divide it in half, and proclaim that this is the true admixture of the population (if, of course, no Negroid Y-chromosomes were found). This seemed, strangely, simultaneously logical and ludicrous, and it would conflict with the results of the above formula, wouldn't it?

Also, do you offhand know what the male Negroid markers are? I believe, but am not sure, that HG8, A, and E*(xE3b) are sub-Saharan.

Hi Frank:

Thanks for your answer. However, now I wonder how much genetic distance exists between Mongolian and Siberian populations. Aren't they closely related as well.

What amazed me is that people of those Asian regions not only resemble physically Native Americans but also have several cultural things that seem related, like shamanism, tipis, canoes, religion, music, etc. I wonder...

Regards,

Omar Vega

It is not that I lack faith in Syke's work. Findings are findings. If one percent of a random 10,000-person sample of Brits have L mtDNA haplotypes, then we can be 95/100 certain that the true recent African admixture in the British population is between 0.86% and 1.14%. My problem was with the sensationalist tone of saying that "one percent of Brits have African or Asian ancestry." Hell, if you go back far enough, everyone (including 100% of Brits) has African or Asian ancestry.

There is a big difference between "admixture" (measurable DNA) and "ancestry." This is because each person can carry only one person's quota of DNA. You had a million ancestors in 1500 but you have only one person's worth of DNA. So 999,999 millionths of your ancestral DNA was discarded long before you were a gleam in your father's eye. Only one/one-millionth of your ancestral DNA was passed down to you. Anyone can measure your admixture. No one can measure the other 999,999 millionths of your ancestry.


This goes to the randomness of the sample. So far, we have assumed that the mtDNA sock is a fair (random) sample of all the other socks in all of the drawers. But in many places and times associated with the slave trade, mixing was mostly via African females and nonAfrican males, rather than vice-versa. And so some people like to adjust for this. If you assume that African male with nonAfrican female mixing never happened, then you can cut the mtDNA number in half; if you assume that it was half as frequent, then you would cut the mtDNA number by one-third, etc. It would be better, of course, to sample both the mtDNA and the SR-Y. Best yet, would be to use the hundreds of newly discovered autosomal markers.


Yes. Also, "B" (Biaka and Mbuti pygmies, Zulus, and !Kung). The two best sources for your answer are "The Phylogeography of Y Chromosome Binary Haplotypes and the Origins of Modern human populations" and "Y Chromosome Sequence Variation and the History of Human Populations" at http://backintyme.com/admixture/underhill01.pdf and http://backintyme.com/admixture/underhill02.pdf, respectively. Incidentally, both of these are by Peter Underhill of Stanford (Cavalli-Sforza's group), the leading expert on SR-Y. So anything by him is worth reading.

But SR-Y nomenclature is not yet as standardized as mtDNA. And so, in order to translate between papers, you need to download the chart titled "A Nomenclature System for the Tree of Human Y-Chromosomal Binary Haplogroups" from http://backintyme.com/admixture/Y-nomenclature.pdf. It also makes a great colorful wall poster!

Not in their mtDNA. Recall that Amerinds are A, B, C, and D (and nothing else). The problem is that some Siberian populations lack either A, B, C, or D. If they are related to Mongolians or Amerinds, there where did the missing haplotypes go? Other Siberians have A, B, C, and D but also have bunches of other stuff. If they are related, then where did the other stuff come from? Specifically:
Nganasans have 22% "other."
Evenks have no B.
Sel’kups are 35% C and 65% other.
Yukagirs have no A or B.
Chukchi have no B and have 29% other.
Siberian Eskimos (Aleuts) have no B or C.
Evens have no A or B and have 35% other.
Koryak have no B and have 46% other.
Nivkhs have 28% D and 72% other.
Udegeys have 18% C and 82% other.

Download Dr. Mullligan's paper. You will find it interesting.