Major factors of protein evolution revealed by eigenvalue decomposition analysis

摘要

Here we attempt to characterize protein evolution by its dominant factors. These factors are revealed by top eigenvectors in the spectrums of eigenvalue decomposition analysis. To reduce the bias induced by closely related sequences in the database, we introduce a parameter, sequence identity by which proteins with sequence identity below certain level are involved in analysis. It is found that, with drop of sequence identity level, residue feature mainly conserved in mutation changes from hydrophobicity to volume. The transition point is at sequence identity ≈45%. As residue hydrophobicity no longer governs residue substitution, it is a doubt whether importance of hydrophobic interaction declines in conserving the family representative properties among remote homologues. So, we also investigate the contribution of hydrophobic interaction in near and remote homologues. In aligned homologues, hydrophobic interaction systems inbuilt in these proteins are aligned too; and can be deemed to be similar and substitutable with each other. With a comparison of aligned hydrophobic interaction systems, we can catch the representative character of hydrophobic interaction for the corresponding protein family. Here top weighted feature in the substitution of hydrophobic interaction systems is revealed as a function of sequence identity. It is found that a shift happens to the type of physical quantity which governs the substitution of hydrophobic interaction. The number of hydrophobic residue is the dominantly unchangeable part in aligned hydrophobic interactions as sequence identity >30%. Below this point, state of internal hydrophobic force which characterizes the residue-residue pairwise interaction is primarily conserved. With view of this shift, intrinsic requirement of protein evolution is sought in the discussion section.