TripletRes

Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks. A contact map prediction based on a triplet of coevolutionary features and deep residual neural networks. TripletRes is a method for protein inter-residue contact prediction. For a query sequence, TripletRes starts with the collection of deep multiple sequence alignments (MSAs) through whole-genome and metagenome sequence databases. Next, three complimentary coevolutionay feature matrices (covariance martrix, precision matrix and the pseudolikelihood maximization) extracted from the MSAs are used to create contact-map models through deep residual convolutional neural network training.