Detecting Convergence of Amino Acid Physicochemical Properties Underlying the Organismal Adaptive Convergent Evolution.

Many studies have proposed various comparative genomic methods to probe the molecular basis for adaptive functional convergence between species, conventionally by detecting the convergence of amino acid states between orthologous protein sequences of these species or lineages. However, different amino acids with similar physicochemical properties at a site may contribute to the functional similarity of the protein. Hence, could the convergence of amino acid physicochemical properties, in addition to state convergence, also contribute to adaptive convergence of organismal functions? Here we grouped amino acids into physicochemically similar classes, and developed computational pipelines to detect the Convergence of Amino Acid Properties (CAAP, https://github.com/shanschen33/CAAP) by modifying previous state convergence detection methods. Investigating three organismal convergence cases including echolocating mammals, marine mammals and woody mangroves, we found genes with CAAP that likely contribute to the respective functional adaptation, supported by orthogonal evidence such as functional enrichment and positive selection analyses. Our findings in multiple cases corroborate the hypothesis that CAAP may underlie adaptive convergent evolution of organismal functions, emphasising the importance of considering sequence features more complex than amino acid states when studying adaptive sequence convergence.