Hacker News Reader

Race and IQ are not scientifically precise tools.

Genetic Structure, Self-Identified Race/Ethnicity, and Confounding in Case-Control Association Studies

We have analyzed genetic data for 326 microsatellite markers that were typed uniformly in a large multiethnic population-based sample of individuals as part of a study of the genetics of hypertension (Family Blood Pressure Program). Subjects identified themselves as belonging to one of four major racial/ethnic groups (white, African American, East Asian, and Hispanic) and were recruited from 15 different geographic locales within the United States and Taiwan. Genetic cluster analysis of the microsatellite markers produced four major clusters, which showed near-perfect correspondence with the four self-reported race/ethnicity categories.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1196372/

So are you saying that if we did the same analysis across Africa, you'd find the same markers throughout the populations in "near-perfect correspondence" for people self-identifying racially as African?

Or is this specific to the populations you studied in the United States?

Dude, I didn't do the study.

Re: Sub Saharan Africans you can get a decent idea if someone has such ancestry from one allele.

As one example, the Duffy Null allele (FY0) has a frequency of almost 100% of Sub-Saharan Africans, but occurs very infrequently in populations outside of this region. A person having this gene is thus more likely to have Sub-Saharan African ancestors.*

http://en.wikipedia.org/wiki/Ancestry-informative_marker

Obviously the more such alleles you test for the better you'll be able to distinguish the populations.

This page gives links to ~300 AIMs which allow you to distinguish ancestry pretty well within Europe.

http://www.snpedia.com/index.php/Ancestry_Informative_Marker

Given that that's possible distinguishing continental ancestry seems to be a problem of application not theory.

One of the co-authors of that paper has withdrawn some of the strong conclusions once based on it in his subsequent co-authored article in the latest edition of Vogel and Motulsky human genetics textbook (cited in my link to a Wikipedia user page in another comment in this thread). Anyway, what the article shows in relation to this thread is not the point the participant above was relying on. It is occasionally possible, with a sample of genetic information about genes not under strong selection pressure, to make a reasonably accurate rough guess of the "race" categorization of many individuals. (I have read the article, and I see that that is the claim, as in the excerpt you kindly quote, but the article also notes exceptions in the sample it examined who were miscategorized.) But those study authors, and all competent human geneticists, do NOT argue the converse: that if you know an individual's race categorization, you know to a reasonable degree of accuracy what specific genes that person has of interest to human behaviors. In fact, we generally do not know that at all about any individual. There are still few genes significant behavorial effect that have been even tentatively identified in a first sample, and the widely replicated result is that ANY "race" group defined by the varying social definitions around the world will still be highly diverse as to any genetic characteristic of interest.

>It is occasionally possible, with a sample of genetic information about genes not under strong selection pressure, to make a reasonably accurate rough guess of the "race" categorization of many individuals. (I have read the article, and I see that that is the claim, as in the excerpt you quote, but the article also notes exceptions in the sample it examined who were miscategorized.) But those study authors, and all competent human geneticists, do NOT argue the converse: that if you know an individual's race categorization, you know to a reasonable degree of accuracy what specific genes that person has of interest to human behaviors.

I don't think the comment you're replying to purported to argue that in the first place. He was just trying to refute the notion that race is merely a matter of social labels; the fact is those labels refer to something real in terms of a person's ancestry that can be detected by genetic tests.

Like you say, it has never been demonstrated that racial groups differ in their genetic propensity for intelligence, but it has also never been demonstrated that they don't. Given the observed IQ differences between racial groups and the fact that these groups differ genetically, a genetic basis for these differences seems like a reasonable thing to hypothesize about, and apply scientific methods of inquiry into. This is why I have a problem with a statement like "race and IQ is not a legitimate matter for scientific inquiry."

This is why I have a problem with a statement like "race and IQ is not a legitimate matter for scientific inquiry."

But both are social-political constructs. They are labels. What do you gain by correlating some feature of human behavior under examination with them? (I provide one answer in another comment: the effective distribution of useful information through existing social channels.)

Yes, racial labels refer to something real. But it's like correlating geological data with the state of California. California refers to something real. But as a scientist why would I want to correlate seismic activity with the state borders of California as opposed to, say, some fault-line that runs out into the Pacific or up into Oregon?

[T]he widely replicated result is that ANY "race" group defined by the varying social definitions around the world will still be highly diverse as to any genetic characteristic of interest.

True, but I can't recall seeing anyone arguing the opposite. And the fact that two distributions overlap doesn't mean that knowing which distribution an individual belongs to contains no information. Men really are on average taller than women.

Any statement which manages to leave out a reasonable self-identification that about 1/6 of the world's population - Asian Indian - would use, is a little suspect.

I'm sure they're working on it. :)

Google Bamshad and Wooding. There are plenty of papers that analyze Indians and show that you can probabilistically identify not just subcontinental ancestry, but caste as well from the genome.

Do you think that a person with measured IQ of 70 is going to have just as foo (e.g. "easy a life", "much success", etc.) as a person with measured IQ of 130?

Your overall position sounds to me like you want crisp, 1-bit, binary, black-or-white definitions or measures for everything, which is a pretty absurd demand. The real-world is n-bit. If you're looking for mathematically rigorous tools, fuzzy logic and probability theory help a lot in this area.

The race problem isn't that hard qualitatively, either: part of the reason we say Obama is a black man is because he says he is himself, another reason is because his skin is more on the darker side of the human skin scale, without prior tanning, than others. You can come up with other reasons I'm sure. You can also go down to the genetic context if you wish, but I'm sure you know the public has been beaten over the head enough times for the past however many decades that race in a genetic context isn't quite as visible as in a visual or cultural context. (Depending on your samples, anyway.)

I don't want 1-bit binaries. In fact, I want acknowledgement of the n-dimensional n-bit nature of the phenomena we call human beings. IQ and race both are trying to fit classifications that properly involve kilobytes if not megabytes of data into convenient oversimplifications that comfortably sit in a single byte.

But in a study comparing disease rates among ethnic groups, African-Americans and Caucasians are treated as disjoint sets.

Are you saying this isn't the case? I'm very curious, as I've heard before that certain diseases have a higher chance of occurring among different races (e.g. blacks are considered more at risk for gout, I think). Is that incorrect?

If I took thousands of people who self-identified as white, and thousands who self-identified as black and I looked at the occurrence of gout in those populations: what would I see?

My understanding is I'd see more gout in the black population. If so, wouldn't that somehow counter your statement that "in a genetic context, most African-Americans are just as much white as black"? Or perhaps, while most African-Americans are like that, some are in fact genetically more "black" in some sense, and they are the ones at a higher risk of gout? I don't really understand.

The question is: what do you gain by categorizing them as self-identified black or white as opposed to not categorizing them at all or categorizing them, say, by the presence of some gene cluster or genetic marker associated with gout?

There is something to be gained, from a social or political angle. A public health advocate in government might say, "Hey, if we see this is a more common problem among a particular social community, maybe we can use the pre-existing channels of information distribution within that community to help them out."

But if your study were not approaching it from this public health angle, then what's the point of correlating gout with this other quasi-scientific socio-political factor? There may be other more suitable categorizations where this particular African-American/Caucasian distinction breaks down.

The question is: what do you gain by categorizing them as self-identified black or white as opposed to not categorizing them at all or categorizing them, say, by the presence of some gene cluster or genetic marker associated with gout?

You gain many things.

If the True outcome is more common in group A than group B, you get a prior (in the Bayesian sense) which can be measured at low cost.

If you are attempting to identify the aforementioned genetic marker, you get to rule out genetic markers which are equally prevalent in both groups A and B. This narrows the search space considerably.

Throwing information (even imperfect information) away is stupid.

Throwing information (even imperfect information) away is stupid.

A falsifier to this proposition using the original question:

If I took thousands of people who self-identified as wearing white sneakers, and thousands who self-identified as wearing black sneakers and I looked at the occurrence of gout in those populations: what would I see?

Obviously one expects a stronger correlation with race based on the fact that members of a racial group are assumed to breed among themselves more frequently and concentrate relevant genetic factors -- if that's what we're looking for. But I suppose that's the point I'm trying to make: let's be clear what we're looking for.

Similarly for cultural factors like diet. But then again, why not cut straight to the dietary data?

In this case, you only get to rule out the genetic markers that are prevalent in both groups once you've established that they are indeed prevalent in both groups. And even then you're running a risk of false correlations for specific individuals because you started with a categorization freighted with a lot of non-scientific baggage.

If I took thousands of people who self-identified as wearing white sneakers, and thousands who self-identified as wearing black sneakers and I looked at the occurrence of gout in those populations: what would I see?

It is unlikely that you would observe a difference between gout incidence between the two groups, and the prior will give no information.

Similarly for cultural factors like diet. But then again, why not cut straight to the dietary data?

Cost. It's very easy for a doctor to ask about race and diet. Genetic markers require moderately expensive lab tests and repeat doctor visits.

Similarly, if you can use simple mechanical prodding to diagnose an injury with a reasonable degree of certainty, it might be worthwhile to skip the expensive MRI.

Further, the dietary or genetic marker associated with a particular diagnosis may be unknown and racial information may be the best available predictor.

If you intern at 23andMe, it's a "hello world" level problem to correlate genetics and ancestry/race. Or just submit a blinded set of saliva samples and see whether their ancestry painting algorithm correctly identifies the pre-1492 ancestral locations of your four grandparents.

This is about real raw data.