Welcome to GWProt’s documentation
GWProt is a computational framework that leverages recent advances in metric geometry, specifically the use of Gromov-Wasserstein (GW) correspondences, for protein structure alignment. GW correspondences find an optimal structural alignment between pairs of proteins by minimizing the distortion of intra-molecular distances among paired residues. GWProt enables the incorporation of biochemical information into structural comparisons and introduces the concept of local geometric distortion, a measure that captures fine-scale conformational differences. Using this framework, we can identify conformational switches in individual proteins, detect functional domains shared among evolutionarily distant proteins, reveal topological rearrangements in homologous folds, or uncover recurrent short structural motifs underlying functional domains. GWProt has been developed by the Cámara Lab. Installation instructions can be found at the GWProt GitHub repository.
Background
Examples