Saturday, April 26, 2014

Small effects at many genes, but are the effects non-additive?


 
Allele dominance (source). A single copy of a dominant allele is as effective as two copies of a recessive allele. The current thinking is that intellectual capacity has increased in humans through new alleles that cause small positive effects at a large number of gene loci. It now seems that some of these new alleles display non-additive effects.
 

How has intellectual capacity increased in the course of human evolution? The current thinking is that natural selection has favored new alleles that cause small positive effects at a large number of gene loci. Over the human genome, these little effects have added up to produce a large effect that distinguishes us from our predecessors.

But are these effects simply additive? Many alleles are dominant, i.e., a single copy has the same effect as two copies. Many alleles also interact with alleles at other gene loci. It would be strange if none of the many gene loci involved in intellectual capacity showed no dominance or interaction. This point was made over a decade ago:

The search for genes associated with variation in IQ will be made more difficult, to the extent that genetic effects on IQ are not additive. We used earlier the illustrative possibility that IQ was affected by 25 genes, each with an equal, additive effect (paragraph 7.15). But some genetic effects, dominance and epistasis, are not additive.

[...] For example, it might be the case that allele 5 of the IGF2R gene is associated with high IQ only if it is accompanied by particular alleles at other loci. In their absence, it is accompanied by normal or even low IQ. If that were true, it would clearly be difficult to detect, and replicate, substantial effects.

[...] Is the genetic variance underlying variation in IQ mostly additive? We noted in Chapter 4 that much research in behavioural genetics assumes this to be the case. But two relatively sophisticated attempts to model IQ variation, while both concluding that the overall broadsense heritability of IQ is about 0.50, also argue that additive genetic variance accounted for no more than about 30% of the overall variation in IQ, while non-additive effects accounted for some 20%. (Nuffield Council on Bioethics, 2002)

Yet many researchers still argue for a simple additive model. Davies et al. (2011) estimated additive genetic variance at 40-51%. As some of the same authors later pointed out, however, the methodology of that study (genome-wide complex trait analysis) ignores non-additive effects: "GCTA estimates additive genetic influence only, so that non-additive effects (gene–gene and gene-environment interaction) are not captured either" (Trzaskowski et al., 2013).
 

Differences among human populations

Davide Piffer (2013) has studied geographic variation in alleles that influence intellectual capacity. He began with seven genes (SNPs) whose different alleles are associated with differences in performance on PISA or IQ tests. Then, for fifty human populations, he looked up the prevalence of each allele that seems to increase performance. Finally, for each population, he calculated the average prevalence of these alleles at all seven genes.

The average prevalence was 39% among East Asians, 36% among Europeans, 32% among Amerindians, 24% among Melanesians and Papuan-New Guineans, and 16% among sub-Saharan Africans. The lowest scores were among San Bushmen (6%) and Mbuti Pygmies (5%). A related finding is that all but one of the alleles seem to be specific to humans and not shared with ancestral primates.

Davide Piffer has now used these geographic differences in allele frequencies to estimate the corresponding geographic differences in “genotypic IQ”, i.e., the genetic component of intellectual capacity:

I had already estimated the African genotypic IQ from my principal component scores extracted from allele frequencies (Piffer, 2013) for different populations. If we take the factor score of people living in equal environmental conditions (Europeans and Japanese), we can figure out how many IQ points each unit score corresponds to. The factor score of Europeans is 0, that of the Japanese is 1.23. The average IQ of Europeans is 99 and that of the Japanese is 105. Thus, 6 IQ points equal a difference of 1.23 factor scores. The factor score of sub-Saharan Africans is -1.73, which is 1.41 times greater than the difference between Europeans and East Asians. Thus, the genotypic IQ difference between Africans and Europeans must be 6*1.41= 8.46. Thus the real African genotypic IQ is 99-8.46= 90.54 (source)

This estimate of 91 seems to contradict the IQ literature, although there is still disagreement over the mean IQ of sub-Saharan Africans. In their review of the literature, Wicherts et al. (2010) argue for a mean of 82, whereas Lynn (2010) puts it at 66. Rindermann (2013) favors a “best guess” of 75. There is some fudging in all of these estimates, since no one really knows how much adjustment should be made for the Flynn Effect. Indeed, what is the potential for IQ gains in societies that are still becoming familiar not only with test taking but also with the entire paradigm of giving standardized answers to standardized questions?

We are on firmer ground when estimating the mean IQ of African Americans, which seems to be around 85, i.e., 15 points below the Euro-American mean. We can argue back and forth over the cause, but the same gap comes up time and again, even when black and white children are adopted into the same home environment. This was the finding of the Minnesota Transracial Adoption Study: a longitudinal study of black, biracial, and white children adopted into white middle-class Minnesotan families, as well as the biological children of the same families (Levin, 1994; Lynn, 1994; Scarr and Weinberg, 1976; Weinberg, Scarr, and Waldman, 1992). IQ was measured when the adopted children were on average 7 years old and the biological children on average 10 years old. They were tested again ten years later. Between the two tests, all four groups declined in mean IQ. On both tests, however, the differences among the four groups remained unchanged, particularly the 15-point gap between blacks and whites. Whatever the cause, it must happen very early in life. Could it be in the womb? We would then have to explain the consistently halfway scores of the biracial children, who were born overwhelmingly to white mothers.

In any case, whether we accept the African American mean of 85 (which is influenced by some admixture with other groups) or the upper estimate of 82 for sub-Saharan Africans, we are still well below the “genotypic” estimate of 91.  Is this an indication of non-additive effects? Do some of the intelligence-boosting alleles display partial dominance? Do some of them interact with other such alleles?
 

References

Davies, G., A. Tenesa, A. Payton, J. Yang, S.E. Harris, D. Liewald, X. Ke., et al. (2011). Genome-wide association studies establish that human intelligence is highly heritable and polygenic, Molecular Psychiatry, 16, 996–1005.
http://www.nature.com/mp/journal/v16/n10/abs/mp201185a.html  

Levin, M. (1994). Comment on the Minnesota transracial adoption study, Intelligence, 19, 13-20.
http://www.sciencedirect.com/science/article/pii/0160289694900493  

Lynn, R. (1994). Some reinterpretations of the Minnesota Transracial Adoption Study, Intelligence, 19, 21-27.
http://www.sciencedirect.com/science/article/pii/0160289694900507  

Lynn, R. (2010). The average IQ of sub-Saharan Africans assessed by the Progressive Matrices: A reply to Wicherts, Dolan, Carlson & van der Maas, Learning and Individual Differences, 20, 152-154.
http://www.sciencedirect.com/science/article/pii/S1041608010000348  

Nuffield Council on Bioethics. (2002). Genetics and human behaviour: The ethical context. London
http://www.nuffieldbioethics.org/sites/default/files/files/Genetics%20and%20behaviour%20Chapter%207%20-%20Review%20of%20the%20evidence%20intelligence.pdf  

Piffer, D. (2013). Factor analysis of population allele frequencies as a simple, novel method of detecting signals of recent polygenic selection: The example of educational attainment and IQ, Interdisciplinary Bio Central, provisional manuscript
http://www.ibc7.org/article/journal_v.php?sid=312  

Rindermann, H. (2013). African cognitive ability: Research, results, divergences and recommendations, Personality and Individual Differences, 55, 229-233.
http://www.sciencedirect.com/science/article/pii/S0191886912003741  

Scarr, S., and Weinberg, R.A. (1976). IQ test performance of Black children adopted by White families, American Psychologist, 31, 726-739.
http://www.kjplanet.com/amp-31-10-726.pdf   

Trzaskowski, M., O.S.P. Davis, J.C. DeFries, J. Yang,  P.M. Visscher, and R. Plomin. (2013). DNA Evidence for strong genome-wide pleiotropy of cognitive and learning abilities, Behavior Genetics, 43(4), 267–273.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3690183/  

Weinberg, R.A., Scarr, S., and Waldman, I.D. (1992). The Minnesota Transracial Adoption Study: A follow-up of IQ test performance at adolescence, Intelligence, 16, 117-135.
http://www.sciencedirect.com/science/article/pii/016028969290028P  

Wicherts, J.M., C.V. Dolan, and H.L.J. van der Maas. (2010). A systematic literature review of the average IQ of sub-Saharan Africans, Intelligence, 38, 1-20.
http://mathsci.free.fr/survey.pdf

19 comments:

barakobama said...

Off subject but i think you'll be interested Peter.

23andme may have just solved the mystery of the occurrence of brown skin in my family: It's descended of Mesolithic Europeans!!!

Most notable my brown skinned uncle does not have the Ala111Thr and Phe374Leu mutations(like Loschbour and La Brana-1) which are fixated in modern Europeans and are suppose to have the biggest lighting effect on modern European skin.

http://www.theapricity.com/forum/showthread.php?123503-23andme-reveals-I-have-relatives-with-Mesolithic-European-decended-dark-skin

Beyond Anon said...

Wow, Peter, every reference has a link.

Do you know of any works that provide a genetic model of IQ that can account for both the mean and the variance in IQ among populations?

Anonymous said...

Davide Piffer has now used these geographic differences in allele frequencies to estimate the corresponding geographic differences in “genotypic IQ”, i.e., the genetic component of intellectual capacity

I'd be more impressed by seeing his estimate for Mexican Mestizos in the US.

Anon 26/04/2014 09:06 said...

Anyway, actually we can apply this more thoroughly, as he's helpfully given the data in his paper to do so -

As the Middle East as a region lags Europe by -0.19 on his main factor score, in an equal environment, their IQ should lag by be (6/1.23)*0.19=, of 1 IQ point, IQ 99 where Europeans are 100.

Assuming Mexican Mestizos are the midpoint between the average of Pima+Maya and Europeans, they would also lag by -0.465 of the factor score, or IQ 97.7 where Europeans are 100.

Going out further and using the Dai score to estimate the performance of Thais, the Dai are advantaged by 0.87 beyond Europeans on the factor score, so their IQ should be 104.2. The Dai and Thai may be different, so if we average the Dai and Cambodian factor score to estimate Thai, (0.32+0.87)/2 = 0.595, converting to an IQ advantage of 102.9.

XXXXXXXXXXXXX said...

Peter Frost said...
This estimate of 91 seems to contradict the IQ literature, although there is still disagreement over the mean IQ of sub-Saharan Africans. In their review of the literature, Wicherts et al. (2010) argue for a mean of 82, whereas Lynn (2010) puts it at 66. Rindermann (2013) favors a “best guess” of 75.

You are confusing phenotypic IQ with genotypic IQ. In Race Differences in Intelligence, Lynn estimated that sub-Saharan Africans have a phenotypic IQ of 67 and a genotypic IQ of 80, as the deprivations of life common in sub-Saharan Africa are depressing their scores below their genetic potential.

B.B.

Sean said...

High IQ in some studies correlates with variants of IGF2R, and it's supposed to be expressed only when from the maternal chromosome.

Unlikely the genotypic IQ difference between Africans and Europeans is the same as the measured difference between biracial children with white mothers and biracial children with black mothers.

While there will be surprises, this is asking us to believe a little bit too much.

ben10 said...

Where would you put Neanderthals? Probably above the contemporary African sapiens sapiens and way above the African Erectus.

Chuck said...

Hi Pete,

I'm surprised that you commented on this. I was afraid that you would when I saw your discussion over at open psychology. A few points:

(1) Piffer's factors scores are unreliable given the low number of alleles used. Also, most alleles were for educational achievement, which taps into more than g.

(2) Piffer anchors his estimate in the Japanese national IQ, on the assumption that the Japanese phenotypic differences (+5) indexes true genetic differences, while the African estimate (-25, Rindermann) indexes true differences + bias. But one can just as validly use the African phenotypic scores as the anchor and reason that the Japanese score is depressed relative to the genotype.

(3) Instead of trying to anchor allele score using one pair of regions (Europe and East Asia or, worse, Europe and Japan), an alternative is to plot the regression for regional IQs on allele frequency when excluding Africa and then see what regional IQ the African allele score would predict. Wicherts et al. discusses this method with regard to national IQ. When this is done, the African scores is well below 80.

(4) Piffer used dubious immigrant substitute IQs in his paper (immigrant selection is a problem). For example, he used a South Asian IQ of 97 based on some Uk Scores; his regional IQ- allele score correlation can't be taken at face value.

(5) Regarding B/W differences in the U.S., you keep citing the MTRAS but this is weak evidence. The black sample size was only, I believe, 20 (+ 60 some mixed race kids). Malloy and I have some unpublished results that present a more ambiguous picture.

(6) As for mixed race differences in the U.S. I have an unpublished meta-review based on 10 national studies that puts the mixed race difference (one black, one white parent) at 40% of the BW gap. I have been really sick so I haven't been able to finish it. (If you want to write it up, I will finish the computation.)

Generally, I wouldn't make much of allele score until s larger number of alleles are used.




Anonymous said...

A recent survey of ethnic IQ differences in Britain may lend some support to Piffer's contention that the genotypic IQ of blacks is in the low 90s rather than the low-to-mid 80s.

The study - involving thousands of students - found that the black-white IQ gap in Britain was only 8 or 9 points, not 15 as in the US.

Biracial students were measured independent of blacks and found to have a mean IQ barely 3 points below whites.

While it's conceivable that selective immigration had the effect of raising the black British IQ average in comparison to African Americans, cultural dynamics are almost certainly also at work.

Compared to blacks in the UK, African Americans face more peer and mass media pressure to conform to a subculture that undervalues intellectual exertion and overvalues athletics and aggressiveness. This is bound to be a negative influence even on black children adopted into white families and living in heavily white communities - if not from their immediate peers, then at least via the television.

Significantly, the only ethnic groups in the UK who appear to fare as poorly on IQ tests as African Americans are the Roma and Irish Travelers. The former have a mixture of South Asian and European ancestry. The latter are scarcely distinguishable from ordinary Irish. But both groups partake heavily in "oppositional" subcultures that may shave quite a few points off their genetic IQ potential. So, like US blacks, they wind up averaging a full standard deviation behind generic whites.

Anonymous said...

@Sean
"Unlikely the genotypic IQ difference between Africans and Europeans is the same as the measured difference between biracial children with white mothers and biracial children with black mothers."

http://en.wikipedia.org/wiki/Sodium-iodide_symporter

"encoded by the SLC5A5 gene"

"NIS expression in the mammary glands is quite a relevant fact since the regulation of iodide absorption and its presence in the breast milk is the main source of iodine for a newborn."

Iodine is necessary for brain development so women with SLC5A5 (or the east Asian equivalent) may have smarter kids than those who don't.

.

@anon

"The study - involving thousands of students - found that the black-white IQ gap in Britain was only 8 or 9 points, not 15 as in the US."

Britain has a major problem with iodine deficiency.

http://www.bbc.co.uk/news/health-22607161

The US has iodized salt so the white kids may be closer to their genetic potential in the US while those in the UK are below their potential.

Chuck said...

Anonymous Anonymous said... "A recent survey of ethnic IQ differences in Britain may lend some support to Piffer's contention that the genotypic IQ of blacks is in the low 90s rather than the low-to-mid 80s."

I dug up this study myself and posted on it over at my prolish blog -- google: "Occidentalist Partially Falsified". It was then reposted on other blogs. I have looked into the UK differences for a while and only found conflicting and difficult to interpret results. The problem is that there is a paucity of published data.

The scores for the UK sample discussed were based on the "Cognitive Ability Test". The latest version in the U.S. also showed anonymously low B/W differences. See number 57 in my meta-analysis: humanvarietiesdotorg.files.wordpress.com/2013/01/tniredo7.jpg

In fact, the U.S. 2010 B/W CAT difference was no different than the U.K. 2010 CAT difference. Yet on many other tests the U.S. B/W differences stands at 1.0 SD. A possible explanation is that the latest addition of the CAT is less g-loaded. Generally, there has been a trend to reduce group differences by reducing test g-loadings. I don't know. My point is that there is much ambiguity in the data, that one can't rely on even a couple of data points, and that one must be careful with interpretations.

That all said, the U.K. gap is surely lower than the U.S. native B/W one. I did conduct a perfunctory meta-analysis on U.S. differences by generation -- about 10 national samples -- and found that 2nd generation Blacks (from Africa and the West Indies) performed eight tenths as poorly as 3rd+ generation Blacks relative to both 2nd and 3rd+ generation Whites. (Naturally, I couldn't get anyone to write up the results -- so they languish like so many others on my desktop.)

Peter Fros_ said...

Beyond Anon,

Most models used to assume that 25 genes influence intelligence. The number is certainly much greater, and these genes vary not only in their relative effect but also in their interactions with each other. So it would be difficult to create a new model without more and better data.

Anon,

"Mexican" is a cultural category. Genetically, the term is meaningless. Romney is of Mexican origin.

XXX,

I'm skeptical. There are areas of malnutrition in sub-Saharan Africa, but most people are well nourished. Also, the effects of malnutrition on IQ are debatable. The severe famine in the Netherlands during 1944-45 had no lasting effects on IQ.

Sean,

No, I don't believe it makes much difference either. But this is an argument that some people raise.

Neanderthals,

Neanderthal intelligence seems to have been much more domain-specific, like a Swiss army knife. Their technology also showed a strange consistency that we don't see in Upper Paleolithic technology. I don't think they were fully self-conscious in the way that we are.

Chuck,

I more or less agree with what you say. Once we get similar data from at least 30 genes or so, we'll be on firmer ground. But his paper is an important first step.

The MTRAS remains the best trans-racial adoption study in the literature. If better studies have been done, why haven't they been published?

Anon,

"Blacks" are a much more heterogeneous group in Britain than they are in the U.S. In the U.S., most are descended from ancestors in West Africa and to a lesser extent in Central Africa. The white admixture rate is about 20 to 25%.

In Great Britain, "blacks" include recent immigrants from all over sub-Saharan Africa, including Somalis and Ethiopians from the Horn of Africa. Somali IQ seems to be only 5 percentage points lower than the European average.

A more important factor, though, is white admixture. Most UK-born "black" children have one white parent. Yes, they can declare themselves as biracial, but just as many say they are "black" (like Obama in the U.S.). I believe this factor alone accounts for most of the difference between U.S. and U.K. data.

A secondary problem is that the British data tends to be "soft." It often involves things like class attendance and subjective evaluations that are easy to manipulate.

Finally, I don't believe that natural selection has operated over eons of time. Yes, the Irish travellers might indeed have a lower genotypic IQ. I see no theoretical reason why not.

ben10 said...

Neanderthals not fully self-conscious ?
You are a bit harsh to 4% of yourself.

Sean said...

It is very silly to assume the current African (or other third world immigrant) population of Britain is at all typical of their homeland. Paul Collier's book explains that immigrating is a risky high stakes investment for the population of poor countries. The poor in Africa just can't afford to immigrate yet. But a diaspora community shelters the immigrant and lowers the various costs to a point where people without modest assets can migrate. So the relatively capable first waves of immigrants are are the dambusters.

Inevitably these claims that the average Africans have roughly comparable IQs result, and they are going to be accepted, as many seem to do in the above comments.

The coming immigration tsunami, which will be self sustaining and ever accelerating, will be increasingly made up of the typical IQ people in Africa and elsewhere.

Anonymous I said...

"Is this an indication of non-additive effects? Do some of the intelligence-boosting alleles display partial dominance? Do some of them interact with other such alleles?"

Maybe, but Piffer's work doesn't necessarily indicate such effects. The Caucasian value of 99 IQ is high, and the Japanese score is only a single datapoint. Using a more reasonable mean IQ of 96 for Europeans means that 9 IQ points equals 1.23 standard scores, and gives an African genetic IQ of 83 points. This is a few points higher than Lynn's estimation of 80, but not enough to raise any eyebrows.

Incidentally, a much better way to do this would be to enter many data points and find a regression line. Doing so would not only give more accurate answers, but would allow a margin for error to be calculated. The slope of the line in a somewhat noisy dataset - as many people don't seem to realize Lynn & Vanhanen's is - can change significantly depending on which datapoints are used.

Bones and Behaviours said...

Peter, about neanderthal psychology. I understand your point about the relative consistency of the Mousterian (though it is partly an artifact perhaps of preservation bias.)

But its a non-sequitur to jump from saying they were not "fully self-conscious in the way that we are." Can you explain?

Peter Fros_ said...

Bones and Behavior,

Henry Harpending and Greg Cochran covered this point in an "outtake" from their last book:

Our favorite hypothesis is that Neanderthals and other archaic humans had a fundamentally different kind of learning than moderns. One of the enduring puzzles is the near-stasis of tool kits in early humans - as we have said before, the Acheulean hand-axe tradition last for almost a million years and extended from the Cape of Good Hope to Germany, while the Mousterian lasted for a quarter of a million years. Somehow these early humans were capable of transmitting a simple material culture for hundreds of thousands of years with little change. More information was transmitted to the next generation than in chimpanzees, but not as much as in modern humans. At the same time, that information was transmitted with surprisingly high accuracy. This must be the case, since random errors in transmission would have caused changes in those tool traditions, resulting in noticeable variation over space and time – which we do not see.

It looks to us as if toolmaking in those populations was, to some extent, innate: genetically determined. Just as song birds are born with a rough genetic template that constrains what songs are learned, early humans may have been born with genetically determined behavioral tendencies that resulted in certain kinds of tools. Genetic transmission of that information has the characteristics required to explain this pattern of simple, near-static technology, since only a limited amount of information can be acquired through natural selection, while the information that is acquired is transmitted with very high accuracy.

http://isteve.blogspot.ca/2009/01/neanderthals.html

Anonymous said...

What do you base the idea Somalis have an IQ of 95 on?

Anonymous said...

"A more important factor, though, is white admixture. Most UK-born "black" children have one white parent. Yes, they can declare themselves as biracial, but just as many say they are "black" (like Obama in the U.S.). I believe this factor alone accounts for most of the difference between U.S. and U.K. data."

I'm not sure exactly what you're arguing, but if it's what I believe to be, that the reason for the lower gap is due to many UK blacks being mixed race, this is doubtful, and is a problem many, many, many people have long made these kinds of discussions- ignoring assortative mating for IQ, or just assuming mixing occurs between respective group averages. With assortative mating, this has been shown to not be the case in black-white mixtures in the US, wherein whites (who are overwhelmingly women) in black-white couplings have lower IQ's on average. Likewise, social stratification for IQ was quite different in pre-modern times as compared to now, which is why I've always found the whole speculation about the contribution of white ancestry to african-american IQ to be incredibly spurious. This was a point I saw raised in a response, in the 70's I believe, to a paper examining the correlation between skin color and IQ in african americans, and it was impossible to know the average IQ of the whites who contributed to the african-american genepool.