%0 Journal Article %T InDeep: 3D fully convolutional neural networks to assist in silico drug design on protein–protein interactions %+ Bioinformatique structurale - Structural Bioinformatics %+ Centre de Bioinformatique (CBIO) %+ Collège Doctoral %A Mallet, Vincent %A Checa Ruano, Luis %A Moine Franel, Alexandra %A Nilges, Michael %A Druart, Karen %A Bouvier, Guillaume %A Sperandio, Olivier %Z V.M. is recipient of a doctoral fellowship from the INCEPTION project [PIA/ANR-16-CONV-0005] and benefits from support from the CRI through ‘Ecole Doctorale FIRE—Programme Bettencourt’. %< avec comité de lecture %@ 1367-4803 %J Bioinformatics %I Oxford University Press (OUP) %V 38 %N 5 %P 1261 - 1268 %8 2021-12-15 %D 2021 %R 10.1093/bioinformatics/btab849 %M 34908131 %Z Computer Science [cs]Journal articles %X Motivation: Protein-protein interactions (PPIs) are key elements in numerous biological pathways and the subject of a growing number of drug discovery projects including against infectious diseases. Designing drugs on PPI targets remains a difficult task and requires extensive efforts to qualify a given interaction as an eligible target. To this end, besides the evident need to determine the role of PPIs in disease-associated pathways and their experimental characterization as therapeutics targets, prediction of their capacity to be bound by other protein partners or modulated by future drugs is of primary importance. Results: We present InDeep, a tool for predicting functional binding sites within proteins that could either host protein epitopes or future drugs. Leveraging deep learning on a curated dataset of PPIs, this tool can proceed to enhanced functional binding site predictions either on experimental structures or along molecular dynamics trajectories. The benchmark of InDeep demonstrates that our tool outperforms state-of-the-art ligandable binding sites predictors when assessing PPI targets but also conventional targets. This offers new opportunities to assist drug design projects on PPIs by identifying pertinent binding pockets at or in the vicinity of PPI interfaces. %G English %Z We thank IBM for their sponsorship for the optimization of the hyper-parameters. In particular, we thank Jean Armand Broyelle, Maxime Deloche and Xavier Vasquez for their expert support. We are grateful to the Institut Pasteur and the CNRS for their continued support for our research. We also thank Arnaud Blondel for feedback and discussions. %2 https://pasteur.hal.science/pasteur-04102799/document %2 https://pasteur.hal.science/pasteur-04102799/file/btab849.pdf %L pasteur-04102799 %U https://pasteur.hal.science/pasteur-04102799 %~ PASTEUR %~ ENSMP %~ CNRS %~ ENSMP_CBIO %~ PARISTECH %~ PSL %~ ENSMP_DEP_MS %~ ENSMP_DR %~ SORBONNE-UNIVERSITE %~ SORBONNE-UNIV %~ UNIV-PARIS %~ UNIVERSITE-PARIS %~ ENSMP-PSL %~ SU-TI %~ ANR %~ ALLIANCE-SU