Positive selection in the human protein-protein interaction network

PIERRE LUISI; DAVID ALVAREZ-PONCE; MARC PYBUS; MARIO A. FARES; JAUME BERTRANPETIT; HAFID LAAYOUNI

Barcelona

Simposio; 2nd Symposium of the Department of Experimental and Health Science; 2014

Departamento de Ciencias Experimentales y de la Salud, Universidad Pompeu Fabra

Genes are subject to disparate selective pressures: they evolve under different strengths of purifying selection and are differently affected by positive selection. However, the biological mechanisms underlying such differences have not been fully understood yet. While most of the studies focused on describing the factors that make a protein more or less constrained in its evolution, the determinants of the impact of positive selection are to be discovered. Genes and proteins often function as parts of rather complex networks of interacting molecules. Therefore, understanding evolutionary forces acting on individual genes will likely benefit from considering their position in such systems. Several lines of evidence indicate that the strength of purifying selection acting on genes is affected by their relative position in molecular networks while much less is known for positive selection. Nevertheless, some observations suggest that the impact of positive selection on genes is also affected by the position of their encoded proteins in molecular networks. Analysis of a few individual pathways suggests that positive selection preferentially targets genes that exert a high degree of control over the overall behavior of the pathway, i.e. at privileged positions. On the other hand, it has been found that genes that have evolved under positive selection since the human-chimp split tended to encode proteins acting at the periphery of the human Protein Interaction Network (PIN). This analysis, however, relied only on the human genome and on the at the time incomplete chimpanzee genome, which limited the power of positive selection inferences. The current availability of genomic data (including the complete genomes of several mammals, plus 1000 human genomes) allows to infer the action of positive selection with an unprecedented power. In agreement with previous results, we observed that purifying selection act at the core of the PIN while genes affected by ancient positive selection (as inferred from comparison of 10 mammalian genomes) tend to act at its periphery. However, the opposite trend was observed for positive selection acting on human populations: signatures of recent positive selection are more likely to be observed for genes acting at the centre of the human PIN.