作者: David Talavera , Simon C. Lovell , Simon Whelan
关键词: Phylogenetic tree 、 Structural genomics 、 Coevolution 、 Measure (mathematics) 、 Phylogenetics 、 Molecular evolution 、 Evolutionary biology 、 Biology 、 Mutation rate 、 Protein structure prediction
摘要: Recent developments in the analysis of amino acid covariation are leading to breakthroughs protein structure prediction, design, and prediction interactome. It is assumed that observed patterns caused by molecular coevolution, where substitutions at one site affect evolutionary forces acting neighboring sites. Our theoretical empirical results cast doubt on this assumption. We demonstrate strongest coevolutionary signal a decrease rate unfeasibly long times required produce coordinated substitutions. find covarying mostly found different branches phylogenetic tree, indicating they independent events may or not be attributable coevolution. These observations undermine hypothesis coevolution primary cause signal. In contrast, we pairs residues with tend have low rates, it gives rise Slowly evolving residue disproportionately located protein's core, which explains methods' ability detect close three dimensions. lead us propose "coevolution paradox": The strength changes means so such highly unlikely occur. As modern methods structural genomics, critical recognize their biases limitations.