Three-Dimensional Convolutional Neural Networks and a Cross-Docked Data Set for Structure-Based Drug Design

By Paul G. Francoeur, Tomohide Masuda, Jocelyn Sunseri, Andrew Jia, Richard B. Iovanisci, Ian Snyder, David Ryan Koes. 2020

242 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction (2022, methodology): “Deep learning on 3D voxel images (via 3D CNNs) or interaction graphs (via GNNs) have improved the traditional hand-designed scoring function used in AutoDock Vina, enabling better fine-tuning of predicted docked poses, as well as direct binding affinity prediction from the 3D complex (McNutt et al.,”
GNINA 1.0: molecular docking with deep learning (2021, methodology): “We removed all PDB IDs that were in the training sets of the CNN models [18, 45], leaving 441 and 178 protein-ligand pairs for the redocking and cross-docking sets, respectively.”
A 3D Generative Model for Structure-Based Drug Design (2022, methodology): “Data We use the CrossDocked dataset [7] following [20].”
3D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Prediction (2023, methodology): “An empirical study on the CrossDocked2020 dataset (Francoeur et al., 2020) shows that TargetDiff generates molecules with more realistic 3D structures and better binding energies towards the protein binding sites compared to the baselines.”
Pocket2Mol: Efficient Molecular Sampling Based on 3D Protein Pockets (2022, unknown): “Conditional generative models are developed to model the 3D atomic density distributions within the 3D pocket structures (Masuda et al., 2020; Luo et al., 2021).”
The impact of cross-docked poses on performance of machine learning classifier for protein–ligand binding pose prediction (2021, methodology): “…al. reported a standardized dataset named CrossDocked2020 set with 22.5 million poses generated by docking ligands into multiple similar binding pockets to better mimic the real-world scenarios, and they comprehensively estimated the scoring and docking powers of their grid-based CNN models [42].”
Generating 3D Molecular Structures Conditional on a Receptor Binding Site with Deep Generative Models (2020, methodology): “A close analogy to images for the 3D molecular structures is a grid of atomic densities, and this approach has been used successfully for binding discrimination for virtual screening [10, 11, 12], pose ranking for docking [10, 6], and binding affinity prediction [13, 6, 14, 15].”
MolGenSurvey: A Systematic Survey in Machine Learning Models for Molecule Design (2022, methodology): “CrossDock2020 [64] contains a total of 22,500,000 poses of ligands docked into multiple similar binding pockets across the Protein Data Bank. scPDB [204] is a comprehensive and up-to-date selection of ligandable binding sites of the Protein Data Bank.”