Boosting Docking-Based Virtual Screening with Deep Learning

By Janaina Cruz Pereira, Ernesto Raúl Caffarena, C. D. Santos. 2016

308 citations recorded by the source corpus.

In this work, we propose a deep learning approach to improve docking-based virtual screening. The deep neural network that is introduced, DeepVS, uses the output of a docking program and learns how to extract relevant features from basic data such as atom and residues types obtained from protein-ligand complexes. Our approach introduces the use of atom and amino acid embeddings and implements an effective way of creating distributed vector representations of protein-ligand complexes by modeling the compound as a set of atom contexts that is further processed by a convolutional layer. One of the main advantages of the proposed method is that it does not require feature engineering. We evaluate DeepVS on the Directory of Useful Decoys (DUD), using the output of two docking programs: Autodock Vina1.1.2 and Dock 6.6. Using a strict evaluation with leave-one-out cross-validation, DeepVS outperforms the docking programs, with regard to both AUC ROC and enrichment factor. Moreover, using the output of Autodock Vina1.1.2, DeepVS achieves an AUC ROC of 0.81, which, to the best of our knowledge, is the best AUC reported so far for virtual screening using the 40 receptors from the DUD.

Subfields: Computational Theory and Mathematics, 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 Molecular Biology

Citations in the corpus

Opportunities and obstacles for deep learning in biology and medicine (2017, background): “AtomNet [34] samples multiple ligand poses within the target binding site, and DeepVS [318] and Ragoza et al.”
Learning from Protein Structure with Geometric Vector Perceptrons (2020, background): “Thanks to their importance and difficulty, such problems, which we broadly refer to as learning from structure, have recently developed into an exciting and promising application area for deep learning (Graves et al., 2020; Ingraham et al., 2019; Pereira et al., 2016; Townshend et al., 2019; Won et ”
The rise of deep learning in drug discovery. (2018, unknown): “Other studies utilizing CNNs or DNNs have also been published [83–85].”
Development and evaluation of a deep learning model for protein–ligand binding affinity prediction (2017, unknown): “…, 2015; Kearnes et al. , 2016; Jastrz¸ebski et al. , 2016; Segler et al. , 2017; Olivecrona et al. , 2017; G´omez-Bombarelli et al. , 2017; Lenselink et al. , 2017; Xu et al. , 2017; Ramsundar et al. , 2017; Wallach et al. , 2015; Ragoza et al. , 2016; Pereira et al. , 2016; Gomes et al. , 2017).”
A Structure-Based Drug Discovery Paradigm (2019, methodology): “reported deep-learning-based virtual screening method where they compared 95,316 decoys with 2950 ligands docked on 40 receptors and those ranked by the deep convolutional neural network showed better performance than other docking programs [130].”
Molecular Docking: Shifting Paradigms in Drug Discovery (2019, methodology): “[125] recently reported an approach based on Convolutional Neural Networks (CNN) called “DeepVS”, which learns the features relevant for the binding of a ligand to a target under study, given a set of docking results.”
DeepDock: Enhancing Ligand-protein Interaction Prediction by a Combination of Ligand and Structure Information (2019, background): “, the context of a reference atom “a” is defined by a set of basic attributes extracted from its neighbourhood [21].”
Key Topics in Molecular Docking for Drug Design (2019, methodology): “Deep learning neural networks have recently been applied to pose prediction and ranking [78,173,177,183,196].”
Benchmarking AlphaFold‐enabled molecular docking predictions for antibiotic discovery (2022, methodology): “Con-comitantly, limitations to the development of more accurate docking methods are the use of rigid protein docking in this and other benchmarking studies (Durrant et al , 2013; Pereira et al , 2016) and the scarcity of benchmarking datasets.”
Artificial intelligence in drug discovery: recent advances and future perspectives (2021, unknown): “Early approaches were inspired by the advances of computer vision and image recognition which were mostly driven by convolutional neural networks [76], and ultimately adapted for bioactivity prediction [77–80].”
Progress in molecular docking (2019, background): “[18] applied deep learning approach in virtual screening, which extracts relevant features from molecular docking data to create the distributed vector representations for protein-ligand complexes.”