Eigenvector centrality for characterization of protein allosteric pathways

Allosteric processes are ubiquitous in macromolecules and regulate biochemical information transfer between spatially distant sites. Despite decades of study, allosteric processes remain generally poorly understood at the molecular level. Here, we introduce the eigenvector centrality measure of mutual information to disentangle the complex interplay of amino acid interactions giving rise to allosteric signaling. The analysis of eigenvector centrality is tested in imidazole glycerol phosphate synthase (IGPS), a prototypical V-type allosteric enzyme. The resulting insights allow us to pinpoint key amino acids in terms of their relevance in the allosteric process, suggesting protein-engineering strategies for control of enzymatic activity.