• Corpus ID: 240288665

Learning physics confers pose-sensitivity in structure-based virtual screening

  title={Learning physics confers pose-sensitivity in structure-based virtual screening},
  author={Pawel Gniewek and Bradley Worley and Katelyn Stafford and Henry van den Bedem and Brandon M. Anderson},
In drug discovery, structure-based virtual high-throughput screening (vHTS) campaigns aim to identify bioactive ligands or “hits” for therapeutic protein targets from docked poses at specific binding sites. However, while generally successful at this task, many deep learning methods are known to be insensitive to protein-ligand interactions, decreasing the reliability of hit detection and hindering discovery at novel binding sites. Here, we overcome this limitation by introducing a class of… 

Figures and Tables from this paper

EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction
A novel and fast fine-tuning model that adjusts torsion angles of a ligand’s rotatable bonds based on closed-form global minima of the von Mises angular distance to a given input atomic point cloud, avoiding previous expensive differential evolution strategies for energy minimization.


AtomNet PoseRanker: Enriching Ligand Pose Quality for Dynamic Proteins in Virtual High Throughput Screens
AtomNet PoseRanker, a graph convolutional network trained to identify, and re-rank crystal-like ligand poses from a sampled ensemble of protein conformations and ligand pose, significantly enriched pose quality in docking to cognate and non-cognate receptors of the PDBbind v2019 dataset.
Protein-Ligand Scoring with Convolutional Neural Networks
This work describes convolutional neural network scoring functions that take as input a comprehensive three-dimensional representation of a protein-ligand interaction and finds that the CNN scoring function outperforms the AutoDock Vina scoring function when ranking poses both for pose prediction and virtual screening.
Development and evaluation of a deep learning model for protein-ligand binding affinity prediction
A novel deep neural network estimating the binding affinity of ligand-receptor complexes is developed, using a 3D convolution to produce a feature map of this representation, treating the atoms of both proteins and ligands in the same manner.
PotentialNet for Molecular Property Prediction
The PotentialNet family of graph convolutions, specifically designed for and achieve state-of-the-art performance for protein–ligand binding affinity, is presented and a cross-validation strategy based on structural homology clustering is introduced that can more accurately measure model generalizability, which crucially distinguishes the aims of machine learning for drug discovery from standard machine learning tasks.
Development and evaluation of a deep learning model for protein–ligand binding affinity prediction
A novel deep neural network is developed estimating the binding affinity of ligand‐receptor complexes by utilizing a 3D convolution to produce a feature map of this representation, treating the atoms of both proteins and ligands in the same manner.
Three-Dimensional Convolutional Neural Networks and a Cross-Docked Data Set for Structure-Based Drug Design
This work presents a new dataset for structure-based machine learning, the CrossDocked2020 set, with 22.5 million poses of ligands docked into multiple similar binding pockets across the Protein Data Bank, and performs a comprehensive evaluation of grid-based convolutional neural network models on this dataset.
Predicting Drug-Target Interaction Using a Novel Graph Neural Network with 3D Structure-Embedded Graph Representation
A novel deep learning approach for predicting drug-target interaction using a graph neural network that extracts the graph feature of intermolecular interactions directly from the 3D structural information on the protein-ligand binding pose.
Docking Screens for Novel Ligands Conferring New Biology.
Improvements in libraries, target quality, and methods that have supported these advances, and the open access resources that make docking accessible are summarized.
AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery
AtomNet is introduced, the first structure-based, deep convolutional neural network designed to predict the bioactivity of small molecules for drug discovery applications and it is shown that AtomNet outperforms previous docking approaches on a diverse set of benchmarks by a large margin.
Hidden bias in the DUD-E dataset leads to misleading performance of deep learning in structure-based virtual screening
It is found that superior enrichment efficiency in CNN models can be attributed to the analogue and decoy bias hidden in the DUD-E dataset rather than successful generalization of the pattern of protein-ligand interactions.