Applications of numerical linear algebra to protein structural analysis: the case of methionine-aromatic motifs

Linear algebraic algorithms uncovered a preponderance of protein methionine residues interacting with two or more aromatic residues. The average geometric relationship between transition metals and methionine-aromatic interactions in PDB crystal structures was assessed with a nearest neighbours-like algorithm (“Met-aromatic”) developed for finding and classifying methionine-aromatic interactions. Here, we assumed that a methionine could interact with one to six midpoints between aromatic carbon atoms in any of phenylalanine, tyrosine, and tryptophan; an integer we termed “Order of Interaction.” Serendipitously, an oversight in Met-aromatic led to the discovery of a significant number of interactions of order exceeding VI, suggesting a large number of methionine residues interacting with two or more aromatics. This was termed the “Bridging Interaction.” Herein, the methioninearomatic and bridging interactions are discussed in light of their possible redox roles.