Visualizing the superfamily of metallo-β-lactamases through sequence similarity network neighborhood connectivity analysis

GONZALEZ, JAVIER M.

Congreso; Biofísica en tiempos de COVID-19 : Primeras Jornadas Virtuales SAB 2020; 2020

Sociedad Argentina de Biofísica

Protein sequence similarity networks (SSNs) constitute a convenient approach to analyze large polypeptide sequence datasets, and have been successfully applied to study a number of protein families over the past decade. SSN analysis is herein combined with traditional cladistic and phenetic phylogenetic analysis (respectively based on multiple sequence alignments and all-against-all three-dimensional protein structure comparisons) in order to assist the ancestral reconstruction and integrative revision of the superfamily of metallo-β-lactamases (MBLs). It is shown that only 198 out of 15,292 representative nodes contain at least one experimentally obtained protein structure in the Protein Data Bank or a manually annotated SwissProt entry, that is to say, only 1.3 % of the superfamily has been functionally and/or structurally characterized. Besides, neighborhood connectivity coloring, which measures local network interconnectivity, is introduced for detection of protein families within SSN clusters. This approach provides a clear picture of how many families remain unexplored in the superfamily, while most MBL research is heavily biased towards a few families. Further research is suggested in order to determine the SSN topological properties, which will be instrumental for the improvement of automated sequence annotation methods. Dr. Liisa Holm is acknowledged for her valuable help with the Dali Lite server. This work was partially funded by ANPCyT grant PICT 2017-4590 to J.M.G. Preprint available at: https://doi.org/10.1101/2020.04.16.045138