Monday, May 15, 2006

Super-hybrids!   posted by Razib @ 5/15/2006 08:59:00 PM

Just a thought. The recent data coming out of genomics, and especially the HapMap, is suggesting that there is a lot of selection in Eurasian populations over the last 10,000 years. So, on the one hand we have loci 1-40 in Chinese being selected for, and loci 20-60 in Europeans being selected form. If an appreciable number of 1-20 and 40-60 are dominant, that is, you gain function from only one copy, then European-Chinese crosses could benefit from complementation. Some of these loci will impact wholly different phenotypes, so that hybrids could hoard talents, kind of like "What if Superman & Wonderwoman mated?" But, another issue that I am wondering about is the possibility that hybrdization could increase dosage and so shift upward the median value of a common phenotype. Because the two populations have found different ways of scaling the same adaptive hill the hybrid can snatch both tricks and double their productivity. Of course, a lot of the recent selection probably has some negative side effects (think of Ashkenazi "overclocking"), so hybrids would also possibly collect a host of susceptibilities and minor ailments. I mean, look at how pensive Agnostic looks :)

Addendum: Admixture of two distinctive populations is liable to increase variance as the number of genomic permutations proliferates and loci become more polymorphic. Does that imply that variance in fitness will also increase? (addressing the issue of fitness in the comment boards should be prefaced by what you mean as fitness!)