The Allele Problem

I read that a typical human is heterozygous at about 6 percent of his or her genes. Also, the heterozygosity per base pair of coding DNA is about 3/10,000 at non-silent sites. At silent sites it is about 1/1000. This means that 2 random versions of a gene drawn from the population have a 6 percent chance of being different, and two base pair positions from coding regions have a 3/10,000 chance of differing at non-silent sites. The coding DNA may be about 100,000 genes and 100 million base pairs. 3/10,000 of this at the non-silent sites and 1/1000 at the silent sites makes at least about 30,000 base pairs of difference between two typical individuals in their coding DNA. This will be spread among about 6 percent of the genes, or, 6,000 genes. This means on the average about 5 differences in base pair per pair of alleles of the same gene that are different.

This is actually a problem, because of the way that evolution is supposed to operate. Beneficial mutations are supposed to spread through the population. If this happens, then all but the most fit allele will be eliminated. And according to the theory of evolution this must have happened quite often, because many genes are different between different species. The hemoglobin gene is often different, for example, implying that some mutation spread through an entire species and eliminated all other alleles. How then can we possibly have 2 alleles differing by 5 base pairs, both common in the population? It must be that mutations to one of the alleles spread through the populatinn but did not eliminate the other allele. And this had to happen at least 5 times for each such pair of alleles, a total of about 30,000 times in the course of human evolution! I think this stretches credulity, even if one accepts such mechanisms as balancing selection.

Back to home page.