DiffMaSIF: Surface-based Protein-Protein Docking with Diffusion Models
Résumé
Predicting protein-protein complexes is a central challenge of computational structural biology. Existing state-of-the-art methods rely on co-evolution learned on large amino acid sequence datasets and thus often fall short on both transient and engineered interfaces (which are of particular interest in therapeutic applications) where co-evolutionary signals are absent or minimal. To address this, we introduce DIFFMASIF, a novel score-based diffusion model for rigid protein-protein docking. Instead of sequence-based features, DIFFMASIF uses a protein molecular surfacebased encoder-decoder architecture to effectively learn physical complementarity. The encoder uses learned geometric features extracted from protein surface point clouds. It directly learns binding site complementary through prediction of contact sites as an auxiliary loss, and also allows for specification of known binding sites during inference. It is followed by a decoder predicting rotation and translation via SO(3) diffusion. We show that DIFFMASIF achieves state-of-the-art among deep Llearning methods for rigid body docking, in particular on structurally novel interfaces and low sequence conservation. This provides a significant advance towards accurate modelling of low co-evolution protein interactions and their many practical applications.
Fichier principal
DiffMaSIF_Surface-based_Protein-Protein_Docking_with_Diffusion_Models.pdf (8.72 Mo)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Licence : CC BY NC - Paternité - Pas d'utilisation commerciale
Licence : CC BY NC - Paternité - Pas d'utilisation commerciale