AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery

TL;DR

AtomNet is a novel deep convolutional neural network that leverages structural information to predict small molecule bioactivity, outperforming traditional docking methods and enabling discovery for targets without known modulators.

Contribution

This paper introduces AtomNet, the first structure-based deep convolutional neural network for bioactivity prediction in drug discovery.

Findings

AtomNet achieves an AUC > 0.9 on 57.8% of DUDE benchmark targets.

It successfully predicts active molecules for targets with no prior modulators.

AtomNet outperforms previous docking approaches by a large margin.

Abstract

Deep convolutional neural networks comprise a subclass of deep neural networks (DNN) with a constrained architecture that leverages the spatial and temporal structure of the domain they model. Convolutional networks achieve the best predictive performance in areas such as speech and image recognition by hierarchically composing simple local features into complex models. Although DNNs have been used in drug discovery for QSAR and ligand-based bioactivity predictions, none of these models have benefited from this powerful convolutional architecture. This paper introduces AtomNet, the first structure-based, deep convolutional neural network designed to predict the bioactivity of small molecules for drug discovery applications. We demonstrate how to apply the convolutional concepts of feature locality and hierarchical composition to the modeling of bioactivity and chemical interactions. In further contrast to existing DNN techniques, we show that AtomNet's application of local convolutional filters to structural target information successfully predicts new active molecules for targets with no previously known modulators. Finally, we show that AtomNet outperforms previous docking approaches on a diverse set of benchmarks by a large margin, achieving an AUC greater than 0.9 on 57.8% of the targets in the DUDE benchmark.