More Recent Comments

Tuesday, December 20, 2016

Is the high frequency of blood type O in native Americans due to random genetic drift?

The frequency of blood type O is very high in some populations of native Americans. In many North American tribes, for example, the frequency is over 90% and often approaches 100%. A majority of individuals in those populations have blood type O (homozygous for the O allele). [see Theme: ABO Blood Types]

Since there's no solid evidence that blood types are adaptive,1 the standard explanation is random genetic drift.

Jerry Coyne explains it in Why Evolution Is True.
One example of evolution by drift may be the unusual frequencies of blood types (as in the ABO system) in the Old Order Amish and Dunker religious communities in America. These are small, isolated, religious groups whose members intermarry—just the right circumstances for rapid evolution by genetic drift.

Accidents of sampling can also happen when a population is founded by just a few immigrants, as occurs when individuals colonize an island or a new area. The almost complete absence of genes producing the B blood type in Native American populations, for example, may reflect the loss of this gene in a small population of humans that colonized North America from Asia around twelve thousand years ago.
Not everyone agrees with this conclusion. There are three possible explanations.
Founder effect
The founding population of the Americas was relatively small so the high frequency of blood type O is just due to chance in a small population. This is classic random genetic drift.

Ancient populations of native Americans had higher frequencies of blood types A and B but they went through a severe population decline in the 1500s and 1600s due to the introduction of diseases from European invaders. More individuals with blood type O survived just by chance. This is similar to the founder effect in that it postulates a connection between allele frequency changes (by drift) and a small population.

The selection model postulates that individuals with blood type O were more likely to survive the diseases introduced by Europeans.
Halverson and Bolnick (2008) realized that the different scenarios could be distinguished by looking at allele frequencies in ancient populations, before extensive contact with Europeans. They looked at the sequences of the two human ABO genes in samples from ancient burial mounds. The ABO gene is located at 9q34.1-q34.2 on chromosome 9. It encodes the enzyme N-acetylgalactosaminyltransferase that's responsible for the ABO blood types. There are many alleles for each blood type [OMIM 110300]. The O alleles are all null alleles (pseudogenes).

Halverson and Bolnick successfully determined the genotypes of 15 individuals who lived about 1800 years ago. Fourteen of them were homozygous for type O alleles (O1 and O1v). The other individual had genotype AO1.

The results show that the high frequency of blood type O was present in populations long before extensive contact with Europeans. This rules out the bottleneck and selection hypotheses and confirms the standard explanation that the effect is due to a founder effect (genetic drift).

It looks like individuals can survive easily without a functional ABO gene. At some point in the future, the null allele might become fixed in the entire population and humans will have one less gene.

1. You will find lots of papers claiming to have found associations between various blood types and different diseases and conditions. Many studies haven't been replicated. There's no consistent pattern.

Halverson, M.S., and Bolnick, D.A. (2008) An ancient DNA test of a founder effect in Native American ABO blood group frequencies. American journal of physical anthropology, 137:342-347. [doi: 10.1002/ajpa.20887]


Jass said...

"Blood group O alleles in Native Americans: implications in the peopling of the Americas.
Estrada-Mena B1, Estrada FJ, Ulloa-Arvizu R, Guido M, Méndez R, Coral R, Canto T, Granados J, Rubí-Castellanos R, Rangel-Villalobos H, García-Carrancá A.
Author information
All major ABO blood alleles are found in most populations worldwide, whereas the majority of Native Americans are nearly exclusively in the O group. O allele molecular characterization could aid in elucidating the possible causes of group O predominance in Native American populations. In this work, we studied exon 6 and 7 sequence diversity in 180 O blood group individuals from four different Mesoamerican populations. Additionally, a comparative analysis of genetic diversity and population structure including South American populations was performed. Results revealed no significant differences among Mesoamerican and South American groups, but showed significant differences within population groups attributable to previously detected differences in genetic drift and founder effects throughout the American continent. Interestingly, in all American populations, the same set of haplotypes O(1), O(1v), and O(1v(G542A)) was present, suggesting the following: (1) that they constitute the main genetic pool of the founding population of the Americas and (2) that they derive from the same ancestral source, partially supporting the single founding population hypothesis. In addition, the consistent and restricted presence of the G542A mutation in Native Americans compared to worldwide populations allows it to be employed as an Ancestry informative marker (AIM). Present knowledge of the peopling of the Americas allows the prediction of the way in which the G542A mutation could have emerged in Beringia, probably during the differentiation process of Asian lineages that gave rise to the founding population of the continent."

Is this what you were looking for Larry?

Unknown said...

Three explanations, but two are essentially the same. A founder effect is a bottleneck.

Robert Byers said...

Genetic drift in this case is welcome to creationism. it rejects natural selection as the mechanism. Something important in biology changes without a selection agenda from the environment.
It also suggests that weird mutations are not needed. the biology change is not from mistakes.
It also hints that envirorment could affect biology without selection needing to eliminate the bulk of the population.
i see this as useful to explain blue eyes/blond hair, red hair etc. these things happening to people groups without just a few getting these traits and the rest selected away. instead all in a tribe suddenly got it.
These blood changes suggest how it was done. A option.

John Harshman said...

The difference is in the date and cause of the bottleneck.

Corneel said...

It also suggests that weird mutations are not needed. the biology change is not from mistakes.

I am sorry Robert, but all O-alleles derive from loss-of-function mutations of the functional gene. Larry said so clearly in the OP:

The O alleles are all null alleles (pseudogenes).

If you follow the link to the OMIM page, you can read for yourself that O-alleles are caused by deletions and point mutations (i.e. "mistakes") in the DNA.

judmarc said...

This seems to be in line with other studies concluding there was a single main wave of migration rather than multiple waves and/or a post-colonization bottleneck. See, e.g., .

Bryan said...

This doesn't quite eliminate the selection model; it only eliminates selection at/around the time of contact with Europeans. Selection at earlier times is still a possibility - but one which should be easily resolvable with a bit of sequencing of the ABO loci.

W. Benson said...

Also, data do not refute a massive Native American population collapse having occurred shortly after European discovery, only that no collapse is evidenced in very fragmentary blood-type (and other types of?) genetic information currently available.

Joe Felsenstein said...

This paper is relevant. They do see asignal of a large population reduction at the time of European contact. It would not have been nearly enough to account fir the fixation of O.

Tim Tyler said...

Malaria is widely thought to be the single strongest selective force affecting blood type - at least among those causes which have been documented. It has long been known that blood group O protects against malaria, and that if you don't have it you are at a greater risk of coma and death. Blood group is quite a variable trait, but it also has a significant effect on the phenotype and frequency dependent selection and heterozygote advantage are alternative explanations for some of that variability. Indeed, one of the common examples of heterozygote advantage involves hemoglobin alleles being affected by parasite frequency.

Larry Moran said...

Thanks, but I was not looking for subsequent articles on the same topic. I just wanted to remind readers that some human phenotypes may not be under selection.

I thought it was best to refer to the original 2008 paper.

Larry Moran said...

The other difference is that allele frequencies can become restricted in populations that remain in the same environment (bottleneck). This is different than the founder effect that involves migration of a small subset of the population.

If the results had indicated high frequencies of the A and B alleles in ancient populations then that would have had severe consequences for the traditional founder effect explanation that's referred to by Jerry in his book.

Larry Moran said...

Remember that drift is the null hypothesis. The onus is on selectionists to provide evidence for selection. At some point we need to stop looking for ways of defending selection because the data is more compatible with drift.

We will never be able to prove the negative (i.e., no selection) but that doesn't mean we have to always point out the obvious; namely, that selection hasn't quite been eliminated.

Larry Moran said...

Correct, and your point is ...?

Larry Moran said...

Some of my ancestors lived in Connecticut in the 1600s. We have plenty of historical records documenting the decline of the native population through war and disease. The settlers were more than happy to assist in any way they could because they got the land.

Several of my male ancestors participated in local raiding parties during the Perquot War. Here's a description from Wikipedia.

The Pequot War was an armed conflict between the Pequot tribe and an alliance of the English colonists of the Massachusetts Bay, Plymouth, and Saybrook colonies and their Native American allies (the Narragansett and Mohegan tribes) which occurred between 1634 and 1638. The Pequots lost the war. At the end, about seven hundred Pequots had been killed or taken into captivity. Hundreds of prisoners were sold into slavery to the West Indies; other survivors were dispersed. The result was the elimination of the Pequot tribe as a viable polity in present-day Southern New England. The colonial authorities classified the tribe as extinct; however, survivors remained in the area and did regain recognition and land along the Thames and Mystic rivers in southeastern Connecticut.

Larry Moran said...

Thank-you for illustrating the adaptationist point of view and how easy it is to spread misinformation.

The founding population of the Americas came from Northern Siberia. That's pretty well established. As far as I know, malaria isn't a major cause of death in Siberia and it was unlikely to be a major cause of death at the end of the ice age.

I expect you want to post several papers that support your adaptationist explanation. Don't bother, I know about them. Not everything published in the scientific literature is correct.

Tim Tyler said...

You apparently accuse me of spreading "misinformation" - yet everything I said was factual and correct.

Bryan said...

I agree, partially (which maybe means I disagree...)

Selection strong enough to fix an allele in a human population over the ~12,000 years since natives came to the Americas should be detectable in the form of higher conservation of DNA flanking the ABO locus than at more distal sites. Thus, with a minimal amount of sequencing you could get a pretty unambiguous answer.

Intellectually, drift is usually the null hypothesis, but that doesn't hold true for actual sequence-based analyses of evolutionary patterns. Many substitution models (at least, those used in protein evolution studies - I'm less familiar with their nucleotide-based brethren) test the null hypothesis of whether a sequence has been positively or negatively selected; deviation from the null indicates either drift, or the other form of selection.

Larry Moran said...

The factually correct statements would be ....

Some publications have shown a correlation between having type O blood and parts of the world where malaria is common. Some publications have claimed to have proof that type O blood protects an individual against malaria.

Other publications dispute these findings and the consensus view these days is that there's no significant selective advantage to having any particular blood type.

Do you see the difference between what I wrote and what you wrote? "Misinformation" can easily be spread by treating a particular science paper as the whole truth without mentioning there are other points of view.

We see that clearly in the debate over junk DNA. It's a serious problem. Just because some scientists say something in a paper doesn't meant it's a "fact." It doesn't become a fact until it has been repeatedly confirmed and accepted by the scientific community.

Larry Moran said...


I know many adaptationists who would be glad to point to selective sweep data as confirmation of their adaptive hypothesis.

Very few of them would be willing to abandon their hypothesis if the data doesn't show evidence of a selective sweep. There's always an excuse. Negative data is never enough.

In this case they have reason to be skeptical of selective sweep date. It doesn't seem to be very accurate in many cases. The data from different labs conflict unless you're dealing with a very few well-confirmed loci.

Tim Tyler said...

From your comment, it doesn't sound as though there is too much actual disagreement here:

When there's frequency-dependent selection at work, at equilibrium, the traits involved typically have equal fitness. If one trait was selectively favored, it would be likely to increase in frequency until a new equilibrium was reached. The observation that there's no net selective advantage to having any particular blood type is compatible with the idea that the traits involved are near neutral and subject to genetic drift - but it is also equally compatible with the idea that they are maintained in the population partly by frequency-dependent selection.

Joe Felsenstein said...

The history, as horrific as it is, still does not have reductions of population size small enough, or for a long enough time, to reduce genetic varability enough to explain the fixation of the O allele by the events of the last 500 years.

For a population to have a bottleneck low enough, and long enough, to massively reduce genetic variability, it must go down to size N and stay there for at least N generations. Thus a bottleneck of 10,000 individuals for 10,000 generations, or 1,000 individuals for 1,000 generations. But with only 500 years (20 generations) for the bottleneck, the South American population would have to have been at a low population size of 20 individuals for that time. So this explanation won't fly. More likely a bottleneck at the time of the original invasion of South America. Or, if other loci do not show similar effects, an effect of selection of some sort.

Larry Moran said...

Thanks Joe. I wasn't implying that the events of 400 years ago could have a serious effect on allele frequencies. Unless, of course, they led to extinction!

I was merely pointing out that the bottleneck was caused by murder as well as disease and the culprits were those upstanding religious immigrants who settled in New England.

I was pretty sure that a recent bottleneck couldn't account for the data based on straightforward population genetics.

What about 15,000 years ago? That's about 500 generations. It still isn't enough time is it? The founding population must have already been considerably enriched in the O allele, right?

Robert Byers said...

ok. I don't get much into genetic matters and will take your word for it.
i'll just keep the idea that selection from pressure in the envirorment is not the origin for the biology change.
So another way for biology change.
This is needed by creationism and who knows how far this could go in explaining biological change.

Robert Byers said...

I understand its been said , and reasonable, that several waves came over. of coarse as a creationist I would say it was only some 3500 years ago for the first Indians to land. yet language grouping shows , probably, a even later migration, came after the first. Then the last one from the Innuit/eskimo etc.
So this would put a wrinkle in any small group concept starting it all with changes in blood.
In short mechanism for biological change has not been observed and so options are there to discover. As a option.

judmarc said...

So the rare blood types would be maintained by negative frequency-dependent selection and the common blood types would be maintained by positive frequency-dependent selection based on what adaptive and maladaptive traits?

Tim Tyler said...

The main idea regarding selection on blood groups is that it is caused by parasites. Parasites tune into existing genotypes, and their spread is hindered by diversity. Blood types are one type of variation that hinders their spread. These dynamics are going to produce negative frequency-dependent selection - where rare types are favored and common types are penalized. If a particular blood type is common that doesn't mean it is enjoying positive frequency-dependent selection. It is more likely that other factors besides frequency-dependent selection are involved. E.g. maybe its favored by selection, maybe it supports a lot of sub-types or maybe there was a population bottleneck and a founder effect.

S Johnson said...

Here's the scoop from the science writer universally acknowledged to be the very best on evolution: Carl Zimmer

Tim Tyler said...

Carl describes the pathogen theory, but he doesn't mention genetic drift. That doesn't seem quite right since one part of the reason why we have the blood types we do is surely accidents of history.

Jim Menegay said...

Two separate questions here: First, why are we polymorphic for ABO, and why has that polymorphism apparently been preserved since we split from the other great apes. Second, why has one particular subgroup of mankind, aboriginal Americans, almost completely lost that polymorphism.

It is entirely possible that the first question has a selective explanation, involving frequency dependent selection and diseases, while the second question (the one Larry is talking about) is explained by neutral processes.

Larry Moran said...

@Jim Manegay

Every human gene is polymorphic. Every gene has multiple alleles. The presence of so much polymorphism in various populations is one of the reasons why Neutral Theory was developed in the 1960s. Extensive polymorphism is easily explained by neutral alleles and random genetic drift but difficult to explain if natural selection acts on every allele.

Frequency dependent selection is a theoretical possibility but in order to be a general explanation it has to apply to just about every nucleotide in our genome. That doesn't seem very likely.

Jim Menegay said...

I agree that it is just as fatuous to invoke frequency dependent selection as a general explanation for polymorphism as it is to invoke selection by default to explain any other ubiquitous feature of the genome. Lewontin and Hubby pointed out the problem, and Kimura solved it.

However, in the case of ABO, the pattern of the polymorphism seems to be conserved in ape and even primate phylogeny. One need not be an ultraselectionist to speculate that there might be a non-neutral explanation for THAT.

Unknown said...

From this discussion it seems to me that the default assumption has to be that the alleles do not give any selected advantage.
The idea ‘there must be some advantage’ certainly can be a useful in getting ones imagination going, but it seems the possible explanations are limitless and unfalsifiable. The assumption of ‘selective advantage’ then leads to too many false positives.

Is that right?

Corneel said...

@Jim Menegay
Yeah, I tried to bring this up a few times as well. The 0-alleles simply occur due to recurring loss-of-function mutations, but the A and B allele represent an ancient polymorphism. If a polymorphism is retained for that long, you definitely want to consider balancing selection.

Graham Jones said...

If you assume the alleles were neutral during the ~200,000 generations since the human-chimp split, there is a puzzle. The human lineage probably had an effective population much less than 200,000 over this time, and under the neutral assumption, you'd expect the alleles to be fixed or lost by now. Adding a simple selection for one or other allele makes fixation even more likely. Hence the attraction of more complex scenarios.

Jass said...

I just wanted to remind readers that some human phenotypes may not be under selection.

Thanks Larry. Do you have an idea what the reason for this phenomenon is?

Larry Moran said...

@Graham Jones
Chimps don't have B alleles and their A and O alleles are different than those in humans. The data suggests that in both lineages the O alleles independently arose by mutation of an A allele. In humans the age of the O alleles is thought to be younger than 3 My.

txpiper said...

"Under selection"

A well-developed illusion. Natural selection is not focused, permeating influence. It is not an ether, or a cosmic exertion. It does not drive or propel, and it sure as hell cannot induce, coerce or bring about any kind of enhancement. All the adjectives dropped in front of the word are meaningless. Nothing is actually being selected. Natural selection is nothing more than organisms ill-suited to their environment dying.

Unknown said...

there is an expected time to fixation, but what is the variance of that expected time?
Just because I expected to throw a double six by now does not mean that because I haven’t anything unusual is taking place.
I think that’s a good analogy.

Oh, I see the point in the case of humans is moot, but my question remains about the variance of the expected time to fixation.
(A reason for me to think the default assumption has to be 'neutral' when it comes to alleles, the time to fixation could be delayed by random luck - one must have actual evidence of selective advantage, it is a mistake to assume it).

This is what I think so far- but I'm a tyro, so I'm sort of guessing.

Faizal Ali said...

Natural selection is nothing more than organisms ill-suited to their environment dying.

Exactly. Now, for the bonus question: If they die without ever reproducing, how many ancestors do they leave to contribute to the gene pool?

Take your time. I'm sure you'll get it if you think really, really hard. There's a good boy.

Tim Tyler said...

Re: "Natural selection is nothing more than organisms ill-suited to their environment dying".

That's more or less what it meant to Darwin. Around the 1930s, the term "natural selection" got redefined to include sexual selection and any other sort of change in gene frequency. The meaning of the term shifted away from "death" to include "reproductive success". For Darwin, sexual selection and natural selection were disjoint. After Darwin's death, natural selection was redefined so that sexual selection became a type of natural selection. Geoffrey Miller covers this redefinition in "The Mating Mind", p.8.

Jmac said...

"Chimps don't have B alleles and their A and O alleles are different than those in humans. The data suggests that in both lineages the O alleles independently arose by mutation of an A allele. In humans the age of the O alleles is thought to be younger than 3 My"


In "the real world of experimental science" that Larry and his science buddies breath, which you know nothing about, things arise and disappear without known causes. All you need to do is to make preconceived conclusions of your experimental evidence, if you have any...that is.

Then, you can make any assumption your heart desires and tell science-fiction stories anyway you want; as long as you stay within the well-walled perimeter known as materialism.

As long as you stay within those boundaries, you can publish anything you want as long as you mention something related to materialism even if it is nonsense.

Welcome to the world of science that has neglected the foundation of its own existence-science.

Joe Felsenstein said...

... and differences in reproductive success are not confined to cases of sexual selection. An individual could achieve higher amounts of reproduction by, for example, being well-fed.

DGA said...

lutesuite, I imagine the ancestors of any unsuited organism are perhaps uncountable, but the descendants are 0.

Ed said...

What a remarkable name, Don Quixote. I imagine you never have read the book?
The story follows the adventures of a hidalgo named Mr. Alonso Quixano who reads so many chivalric romances that he loses his sanity and decides to set out to revive chivalry, undo wrongs, and bring justice to the world, under the name Don Quixote de la Mancha.
Don Quixote, in the first part of the book, does not see the world for what it is and prefers to imagine that he is living out a knightly story.

Your ramblings fit the description of the character perfectly! Well done, an A+ for role playing, unfortunately a E for knowledge.

judmarc said...

By "How many ancestors do they leave," lutesuite was IIUC essentially saying "How many of them *become* ancestors," or as you put it, DGA, how many descendants to they have, to which you gave the (eventual, probable) answer.

judmarc said...

This has been a really excellent, informative thread thus far. Hope it doesn't get sidetracked by the usual know-nothings.

Faizal Ali said...

Thanks for catching the error! Of course, I meant to write "descendants". Hopefully that didn't cause too much confusion for Don Quixote.

txpiper said...

Yes, this is all sufficiently elevated to avoid controversy. Nobody is going to crash and burn when the dispute is between feckless natural selection and nebulous genetic drift. But, shift the discussion to how either of these champions designed your blood distribution system, or the valves in that system, the excellence (and your intense interest in 'science') will disappear.

judmarc said...

Not at all. My intense interest in science leads me to read more about it, and absorb elementary principles such as natural selection *reducing* variation, as does drift, by fixation. So there you are railing on about something you think is a problem, when everyone here is light-years ahead of you in understanding the role it plays, and the fact that it isn't a problem at all.

txpiper said...

Then show some light-years-ahead application, and explain how fabulous and powerful random DNA replication errors, and thundering natural selection ever producd anything.

But that would be a different kind of 'science'. The kind that is not fun at all.

Robert Byers said...

A YEC would add that the purpose of creature groups in reproducing successfully only includes sexual selection. i saw excellernt youtube videos on tasmanian devils and Hyaenas which show the whole group is focused on being strong and ensuring strength in all members. sexual selection is only a special case in this agenda.

judmarc said...

But that would be a different kind of 'science'. The kind that is not fun at all.

To the contrary. I enjoy reading a very few of the literally thousands of detailed peer reviewed papers produced annually on various aspects of these processes. If you don't bother to read, or if you do read, don't bother to develop an understanding, this is not my problem. Since you have in other comments, IIRC, noted that you believe in a young Earth, it's evident there's an awful lot of science you haven't bothered to understand. Usually such a lack of understanding in anyone otherwise reasonably intelligent stems from being motivated *not* to understand, which is what I would venture to guess is the case with you.

anonymous said...

@ Professor Joe Felsenstein

Regarding “horrible history”: the 15th Century was very horrible for European Jewry resulting on a bottle neck of 330 individuals for all Europe. The problem is that such a small population was probably further fragmented into smaller populations isolated from each other. (Fig 4:

The story becomes more complicated when considering all exceptions to the general rules assumed by mathematical modeling. For example, it would appear that Ashkenazim have four major and numerous minor founders according to maternally inherited mitochondrial lineages.

The high incidence of idiopathic torsion dystonia (ITD) among Ashkenazim was examined more than two decades ago by Neil Risch and provided some intriguing insights: The ancestral mutation probably occurred around 1650, in either Lithuania or Belorussia.

However, not all Ashkenazi Jews enjoyed equal reproductive success. Family genealogies suggest that a privileged and relatively affluent few were considered more desirable marriage partners and therefore left a disproportionate contribution to a very small gene pool. In contrast, poorer Jewish families, for a variety of reasons, had a less significant impact on allele frequency in that same small gene pool.

Risch thinks the original ITD mutation occurred by chance in an affluent family, and spread rapidly because affluent Jews tended to intermarry and have larger families.

Your point is still well taken, however. For example, none of many unique Ashkenazi genes are “fixed” even though the prevalence of many deleterious alleles are higher than straight-forward mathematical would predict. At least nobody is talking about Tay-Sachs heterozygosity imparting partial resistance to Tuberculosis. The Quebec data manifesting an independently derived version of Tay-Sachs puts that “just-so” hypothesis to rest.