Deep Learning to Therapeutically Target Unreported Complexes.

Journal: Trends in pharmacological sciences
Published Date:

Abstract

The disruption of large protein-protein (PP) interfaces remains a challenge in targeted therapy. Designing drugs that compete with binding partners is daunting, especially when the structure of the protein complex is unknown. To address the problem we propose a deep protein databank (PDB) learning platform to discover targetable epitopes for complex-disruptive leads.

Authors

  • Ariel Fernández
    National Research Council (CONICET), Rivadavia 1917, Buenos Aires 1033, Argentina; INQUISUR (UNS-CONICET), Avenida Alem 1253, Bahia Blanca 8000, Argentina; AF Innovation Pharma Consultancy GmbH, Avenida del Libertador 1092, Buenos Aires 1112, Argentina; Collegium Basilea, Institute for Advanced Study, Hochstrasse 51, Basel 4053, Switzerland. Electronic address: ariel@afinnovation.com.