A simple tool to explore the distance distribution of correlated mutations in proteins

The analysis of correlated mutations in protein sequence alignments is of considerable interest, since it may provide useful energetic and even structural information (ideally, residue contacts). However, a number of recent experimental studies support the existence of long-distance communication in proteins, a fact that may lead to correlation between distant residues. We introduce in this work a simple statistical procedure to describe the relation structure--alignments on the basis of the residue--residue distance dependence of the number of residue couples over given thresholds of a correlation measure (such as a covariance value). This procedure may lead to clear pictures of the distance distribution of correlated mutations and may provide a simple but efficient tool to explore the different structural features that are reflected in the sequence alignments.