Pattern and process in the molecular evolution of the order primates

The conceptual separation of phylogenetic pattern and evolutionary process is a necessary prerequisite for evolutionary inference. On the assumption of a phylogenetic pattern inferences were made as to the evolutionary processes operating at the molecular level. It was found that the neutral theory provided an inadequate explanation for a number of observed patterns of evolution in the haemoglobin genes of vertebrates. In the light of such a finding a new model of molecular evolution was formulated stressing that neutralist and selectionist approaches cannot be defended in isolation, that every nucleotide in the genome may be under the scrutiny of natural selection, and that more attention must be paid to gene structure rather than simply gene product. DNA turnover was also shown to have played an important part in shaping patterns of molecular evolution. Conversely, assumptions about evolutionary processes are made explicit in probabilistic models of molecular evolution. Probabilistic models are appropriate in that their assumptions are open to test and future refinement, in that they recognise the stochastic nature of the evolutionary process and in that many sequence changes are not the result of a simple process of evolutionary divergence. The further assumption of a stochastically uniform rate of molecular evolution is difficult to defend on empirical grounds (although examples evidently exist), but it serves a useful role in examining the nature of evolutionary processes. Using the probabilistic approach it was found that, on current evidence, it is not possible to resolve either the phylogenetic position of the primates within the eutherian mammals or the interrelationships of the Hominoidea. Important findings were that genes from different genomic regions often supported different phylogenetic relationships as did different genes from within a single functional complex (specially the <IMG WIDTH=9 HEIGHT=24 ALIGN=MIDDLE SRC="/maths/beta.gif"> globins). Suitable caution must be paid to all phylogenies estimated on the basis of molecular sequence data.