Sunday, December 19, 2010

How we are evolving and more on the "missing heritability" problem














Although I suggested  we should skip articles for next lab-meeting (Wednesday 22 December), I changed my mind after finding a couple of interesting papers in the popular science magazine Scientific American. Both deal with human evolution and human genetics, and should be an easy read, and are of much principal interest, also to us.

The first one raises the question "How we are evolving" and it discusses the recent recent data obtained from "Next-generation sequencing" that have elucidated the evolutionary history of coding genes in Homo sapiens. It turns out that so-called "selective sweeps" are quite unusual in our genome, i. e. alleles that have quickly increased in frequency in response to local selection pressures and in local environments. Thus, the case of the recent demonstrated selective sweep in the genes affecting haemoglobin production when humans invaded the Tibetan Plateau that we discussed earlier this year, are atypical, and do not seem to be the norm (according to the article in Scientific American).

Instead, it appears that the global distribution of human coding genes more reflects recent migration histories: genes that increased in frequency at a certain place were then carried along to new places as human colonized the Earth. This implies that selective sweeps are rare, and they mainly happen when there is strong and sustained selection that is consistent over relatively long time at a given place. This implies that selection might in general be relative weak, in humans at least. Perhaps humans are able to adapt (culturally) and counteract selection pressures by cultural adaptations, such as clothing, food habits etc., which weakens natural selection in many environments?

The other article in Scientific American deals with the so-called "missing heritability"-problem, i. e. the uncomfortable facts that there are very few genes explaining a substantial amount of variation in diseases. Thus, most large-effect genes are quite rare in the populations, whereas those genes that are common explain only a tiny fraction (in the order of a few percent) of the disease-related variation. This has caused despair among molecular biologists and biomedical researchers, since the huge amount of sequencing data has not solved many of Homo sapiens biomedical problems (contrary to promises in the most enthusiastic heydays of genome sequencing). In spite of thousands of variable SNP:s ("single-nucleotide polymorphisms"), only a few percent of the total amount of variation in human phenotypic traits are explained by these molecular markers. This can either be interpreted as a great disappointment (in relation to promises), a complete failure for molecular biology or a great triumph of quantitative genetics, but as usual, the questions that are raised by these findings might be more interesting than the exact answers. The solution? Well, certainly not to sequence more or try to obtain more SNP:s, but rather to think harder (surprisingly, an underutilized strategy in much of today's science!).

The two articles can be found here and here, and I think you will enjoy both, as they are relatively easy read. Let's have a brief and informal discussion about these papers on Wednesday, after we've had some glögg!

1 comment:

  1. Long live quant gen! I don't want to have gone through all this pain learning it to have it taken away now ;-)

    Happy Xmas and New Year to one and all. Hope to see some of you in the new year!

    Cheers, Tom

    ReplyDelete