Mapping protein sequence spaces by recurrence quantification analysis: a case study on chimeric structures

Recurrence quantification analysis (RQA) was used to characterize the folding properties of 22 chimeric sequences derived from two parent proteins of similar length but different three-dimensional arrangement. A non-linear relation between sequence data and their RQA representation was revealed, which points to new information carried by this method as compared with classical best-alignment methods. This new information is significantly correlated with the folding properties of the hybrid polypeptide chains, as substantiated by careful statistical analysis of the recurrence plots' numerical descriptors, thus encouraging their systematic use to complement sequence data in both proteomics and protein engineering tasks. Even the direct visual screening of the qualitative graphical features of recurrence plots is shown to provide useful hints to discriminate between different recurrence structures of protein sequences.