Haldane and the first estimates of the human mutation rate

Abstract

Mutation is the ultimate source of genetic no velty. As such, the rate at which new mutations arise is a central i ssue in gene tics, with profound implications for both ev olution and human health. Haldane’s 1935 paper (reprinted in this issue, pages 235–244) was the first to provide a careful estimate of the mutation rate in h umans. In fact, previous studies had already provided est imates of mutation rates in other organism s (Muller 1928; Stadler 1932). The chief difficulty confronting exper imentalists was the fact that mutations occur at a very low rate. Muller (1928) was able to measure mutation rates in Drosophila melanogaster using a scheme of balanced lethals. By workin g with flies, he was also able to breed many individuals and score large numbers of progeny easily. Stadler (1932), working with maize, was able to count millions of seeds to arrive at mutation rates mostly between 10 –4 and 10 –6 for eight different genes. These approaches were clearly not available in humans, and it was recognized that measures of the mutation rate had to come from other methods. Haldane (1927) was the first to develop the fo rmal theory for equilibrium frequencies of alleles in mut ation– selection balance, and these calculations later formed the basis for the indirect estimates of mutation rate in his 1935 paper. Haldane suggested that if selection acts against deleterious alleles it must be balanced by mut ation pressure to generate observ ed allele frequencies in a population. This provided a straightforwar d way to est imate the mutation rate for deleterious alleles: if the fr equency of an allele could be measured and if the strength of selection could be estimated, it should be po ssible to calculate the mutation rate. Since most mutations in genes are harmful, the mutation rate for deleterious a lleles will be only a slight underestimate of the total mut ation rate. For autosomal dominant mutations, the equilibrium allele frequency, q, is simply µ/s, where µ is the mutation rate and s is the selection coefficient associated with the del eterious mutant. This well-known result can now be found in any population genetics textbook (e.g. Crow and K imura