DeepREAL: a deep learning powered multi-scale modeling framework for predicting out-of-distribution ligand-induced GPCR activity

Abstract

Motivation Drug discovery has witnessed intensive exploration of predictive modeling of drug-target physical interactions over two decades. However, a critical knowledge gap needs to be filled for correlating drug-target interactions with clinical outcomes: predicting genome-wide receptor activities or function selectivity, especially agonist vs. antagonist, induced by novel chemicals. Two major obstacles compound the difficulty on this task: known data of receptor activity is far too scarce to train a robust model in light of genome-scale applications, and real-world applications need to deploy a model on data from various shifted distributions. Results To address these challenges, we have developed an end-to-end deep learning framework, DeepREAL, for multi-scale modeling of genome-wide ligand-induced receptor activities. DeepREAL utilizes self-supervised learning on tens of millions of protein sequences and pre-trained binary interaction classification to solve the data distribution shift and data scarcity problems. Extensive benchmark studies on GPCRs, which simulate real-world scenarios, demonstrate that DeepREAL achieves state-of-the-art performances in out-of-distribution settings. DeepREAL can be extended to other gene families beyond GPCRs. Availability All data used are downloaded from Pfam Mistry et al., 2020, GLASS Chan et al., 2015 and IUPHAR/BPS and the data from reference Sakamuru et al., 2021. Readers are directed to their official website for original data. Code is available on GitHub https://github.com/XieResearchGroup/DeepREAL. Supplementary information Supporting figures are available at Bioinformatics online.

Predictive Modeling in Medicine

“…Specifically, there is a need for forecasting genome-wide receptor activities and function selectivity that are brought about by new chemicals-particularly with regard to agonist versus antagonist impacts. However, achieving this objective proves challenging due to insufficient data on receptor activity as well as the necessity of training models with diverse shifted distributions suitable for real-world applications [43].…”

Artificial intelligence in recommender systems

“…To probe dark gene families, Cai et al [2023] propose an innovative OOD meta-learning algorithm PortalCG to generalize from distinct gene families to dark gene family. Because of the challenge of the scarcity of receptor activity data, Cai et al [2022a] propose a self-supervised method DeepREAL to mitigate distribution shifts. To assess the OOD generalization ability of previous drug-target interaction works, Torrisi et al [2022] provide a generalization ability evaluation by including systematic test sample separations.…”

PSICHIC: physicochemical graph neural network for learning protein-ligand interaction fingerprints from sequence data

“…Computational methods for understanding these interactions and predicting associated properties can facilitate cost-effective exploration of vast chemical spaces prior to experimental testing, thereby accelerating drug discovery 1,2 . In response, significant research efforts have been focusing on developing computational methods to accurately predict the binding affinity [3][4][5][6][7][8][9] and functional effects of a small molecule on its protein target [10][11][12][13] .…”

“…Furthermore, unlike structure-based approaches [10][11][12] , sequence-based approaches are not sensitive to the protein structural state when predicting the functional effects of a small molecule on its protein target. Instead, sequence-based approaches can directly ascertain binding affinities and the functional effects of protein-ligand interactions 13 . To this end, deep learning methods applied to sequence data offer increasingly accurate predictions 3,23 .…”