Genetics and the evolutionary process

Population genetics was put forward as a mathematical theory between 1918 a nd 1932 and played a leading part in the rediscovery of the concept of natu ral selection. As an autonomous science developing Mendel's laws at the pop ulation scale and a key element of the Darwinian theory of evolution, its d ual status led its practioners to initially overlook some consequences of M endelism not accounted for by the Darwinian theory including random drift a nd the cost of selection. The latter were put forward on purely theoretical grounds in the 1950s, but their importance was acknowledged only when empi rical data on protein evolution and enzyme polymorphism (since 1965) and on DNA variation (since 1983) were obtained. The neutralist/selectionist deba te that ensued involved disagreement over the scientific method as well as over the mechanisms of molecular evolution. Population genetics has long as sumed the existence of natural selection a priori. It has since recentred a round the null hypothesis that molecular evolution is neutral. This new app roach, applied to sequence comparison and to the study of linkage disequili brium, is logically more justified, yet empirical observa- tions derived fr om it paradoxically show the overwhelming importance of selective effects w ithin genomes. (C) 2000 Academie des sciences/Editions scientifiques et med icales Elsevier SAS.