Graph Convolutional Neural Networks for Predicting Drug-Target Interactions

By Wen Torng, Russ B. Altman. 2018

317 citations recorded by the source corpus.

Accurate determination of target-ligand interactions is crucial in the drug discovery process. In this paper, we propose a two-staged graph-convolutional (Graph-CNN) framework for predicting protein-ligand interactions. We first describe an unsupervised graph-autoencoder to learn fixed-size representations of protein pockets. Two Graph-CNNs are then trained to automatically extract features from pocket graphs and 2D molecular graphs, respectively. We demonstrate that graph-autoencoders can learn meaningful fixed-size representation for protein pockets of varying sizes and the Graph-CNN framework can effectively capture protein-ligand binding interactions without relying on target-ligand co-complexes. Across several metrics, Graph-CNNs achieved better or comparable performance to 3DCNN ligand-scoring, AutoDock Vina, RF-Score, and NNScore on common virtual screening benchmark datasets. Visualization of key pocket residues and ligand atoms contributing to the classification decisions confirms that our networks recognize meaningful interactions between pockets and ligands. Availability and Implementation Contact: rbaltman@stanford.edu Supplementary information:

Subfields: Computational Theory and Mathematics, Materials Chemistry, Molecular Biology

How researchers see this paper

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

In Computational Theory and Mathematics

In Materials Chemistry

In Molecular Biology

Citations in the corpus

EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction (2022, methodology): “Recently, DL changed this paradigm, e.g., using 3D CNNs (Aggarwal et al., 2021; Jim´enez et al., 2017; Torng & Altman, 2019b) or sequence models (Sankararaman et al., 2010).”
Graph Neural Networks and Their Current Applications in Bioinformatics (2021, background): “Torng and Altman (2019) established an unsupervised graph autoencoder to learn the representation of protein pockets without relying on the target–ligand complexes, where features were extracted from the pocket graph and 2D ligand graph by GCN, respectively.”
Protein Docking Model Evaluation by Graph Neural Networks (2020, methodology): “We used the same features as used in previous works (Lim et al., 2019; Torng and Altman, 2019) as shown in Table 2.”
Deep learning in drug discovery: an integrative review and future challenges (2022, background): “In a different DTI investigation study, Torng et al. employed a graph without supervision to become familiar with constant size depictions of protein binding sites (Torng and Altman 2019).”
Machine-learning methods for ligand-protein molecular docking. (2021, methodology): “Torgn et al. propose a VS method that uses two graphs to represent the target and the ligand [95].”
Deep Learning with Graph Convolutional Networks: An Overview and Latest Applications in Computational Intelligence (2023, methodology): “[128] used a graph convolutional neural network to encode atoms, bonds, and distances, which can better utilize the information in the graph structure.”
Geometric deep learning on molecular representations (2021, background): “Since their introduction as a means to predict quantum chemical properties of small molecules at the density functional theory (DFT) level [5], GNNs have found many applications in quantum chemistry [68–72], drug discovery [37, 73, 74], CASP [75], and molecular property prediction [76, 77].”
Utilizing graph machine learning within drug discovery and development (2021, background): “[108] focused on the task of associating drugs with protein pockets they can bind to.”