Saturday, April 11, 2009

Appropriate vs. intrusive genetics I. Natural selection in humans

Human DNA sequences are being scrutinized and mined by many investigators, for anything that can be found in them. The amount and detail of new data now available is unprecedented in history, so the interest is natural. Much can be said about these data, and there are countless papers being written about them. But in this enthusiasm much less is being said about when and whether interpretation becomes over-interpretation, and when these genetic investigations have non-trivial potential to be harmful to the society that is paying for the research. So this is an appropriate place to voice some of the issues, as we see them, at least, and several areas now under intense investigation deserve some attention. This is the first of several posts on this subject.

1. The search for evidence for natural selection in the history of different modern human populations. Humans vary geographically in lots of ways. 'Race' is a traditional term to refer to that variation, and generally based on traits that are visually obvious. However, the term race is properly discredited because it is usually greatly oversimplified, over-categrorical, historically and notoriously difficult to define with any scientific rigor, and has been used to discriminate against people in the worst ways. Skin color is the classical 'race' trait, and certainly varies globally for genetic reasons. But Victorian anthropologists spent a lot of time classifying and identifying additional race-traits. These were often given adaptive explanations, and eugenicists and the Nazis put such explanations into political practice, with tragic consequences. The same traits are still used today to characterize human populations.

As a reaction to World War II, most scientists became restrained in the pursuit of 'race' biology. Some work on racial variation was done with beneficent intent, such as studies of anti-malarial genetic resistance in sickle-cell, and other diseases. Nonetheless, a few mavericks insisted on racial genetic studies, especially in what is known as 'behavior genetics', often not-surprisingly focused on socially sensitive traits, like criminality (defined by the upper class as prohibited lower-class acts). Above all, though, always in the background, or perhaps basement, were studies of race differences in intelligence. That hasn't gone away.

Searches for selection among human groups by definition identify what is 'good' in the sense of having been favored, vs what is 'bad', that which was disfavored by Nature. The fact that all groups still vary greatly, the 'bad' is still here (e.g., in the genes of some inferior 'races'?), and that the overlap between them is usually much greater than their average difference (as in the figure), are important, conveniently overlooked, subtle issues beyond the space of this posting. It's all too easy to say that, genetically, blacks are this, whites are that, or males are this, females are that.

In the current fervor for things genetic, searches for racial differences in various traits has been creeping to respectability again. Often couched in terms of 'disease', which is often a transparent way to rationalize grant support for such studies, or a way to market race-specific medicine, the stand-by racial traits, including visual traits, are again being discussed.

Humans within as well as between groups differ in essentially every trait. Most traits have substantial heritability or familial aggregation, suggesting that genetic variation contributes to them. In principle, it can be important to understand the specific genes involved. When this is done in regard to rabbits, butterflies, or plants nobody cares and good knowledge can come of it (though explanations are often overstated or oversimplified).

But when this same kind of searching is applied to humans, usually from samples collected with prior conceptions of geographic variation in mind (that is, sampling is from the classical 'races' even if subconsciously or never saying so explicitly), it is not just detached science, but it becomes relevant to the societies that are paying for the research. The work can easily become, and often is, related to value judgments about who's good and who's bad, or who's better and who's worse in this or that group, or which group is more advanced than which other group, or where to put supporting or repressive societal effort, such as in racial profiling in medicine, investment in education, or forensic genetics.

So, there is an important issue of what kinds of science should be acceptable or paid for, and what kinds of samples are legitimate--and who should decide. We should not be captives of past history, but history shows what monsters can be loosed by acquiescing in whatever scientists want to do. It is naive or self-interested to pretend that we in the 'civilized' world are above repetitions of past racist disasters--even if they would, if they occurred, take somewhat different form.

If, regardless of intent, there is reasonable potential for claims of findings to lead to discrimination, then it is reasonable to suggest that this kind of work should not be done. There are, after all, all sorts of research that IRBs (Institutional Review Boards, whose job is to oversee university research) don't approve. There are historical precedents for the consequences of looking the other way rather than speaking up and opposing such work, even though science censorship raises problems of its own. And yet we each have our own ideas about what is legitimate, what is downright good, interesting, or great to do, and what should not be done.

In future posts we will discuss how these issues apply directly to forensic genetics, genetic ancestry estimation, and phenotype prediction (of disease, or of what a face looks like, etc.).

No comments: