Improving protein–ligand docking and screening accuracies by incorporating a scoring function correction term

Abstract

Scoring functions are important components in molecular docking for structure-based drug discovery. Traditional scoring functions, generally empirical- or force field-based, are robust and have proven to be useful for identifying hits and lead optimizations. Although multiple highly accurate deep learning- or machine learning-based scoring functions have been developed, their direct applications for docking and screening are limited. We describe a novel strategy to develop a reliable protein–ligand scoring function by augmenting the traditional scoring function Vina score using a correction term (OnionNet-SFCT). The correction term is developed based on an AdaBoost random forest model, utilizing multiple layers of contacts formed between protein residues and ligand atoms. In addition to the Vina score, the model considerably enhances the AutoDock Vina prediction abilities for docking and screening tasks based on different benchmarks (such as cross-docking dataset, CASF-2016, DUD-E and DUD-AD). Furthermore, our model could be combined with multiple docking applications to increase pose selection accuracies and screening abilities, indicating its wide usage for structure-based drug discoveries. Furthermore, in a reverse practice, the combined scoring strategy successfully identified multiple known receptors of a plant hormone. To summarize, the results show that the combination of data-driven model (OnionNet-SFCT) and empirical scoring function (Vina score) is a good scoring strategy that could be useful for structure-based drug discoveries and potentially target fishing in future.

Boosting Protein–Ligand Binding Pose Prediction and Virtual Screening Based on Residue–Atom Distance Likelihood Potential and Graph Transformer

“…Despite this, extensive efforts have been made to develop MLSFs with a wider application domain during the past few years. − In 2017 and 2019, Zhang’s group successively developed Δ Vina RF 20 and Δ Vina XGB, which employ ML algorithms to fit correction terms to AutoDock Vina scores rather than conventionally fit the final binding scores. These two methods could rank the top three in all four tasks (i.e., scoring, docking, ranking, and screening) in the Comparative Assessment of Scoring Functions (CASF) benchmark, although their VS performance in our previous assessment was unsatisfactory .…”

“…This new SF could not only achieve excellent scoring, ranking, and screening powers on the CASF benchmark but also outperform the baselines in terms of the large-scale docking-based VS on the LIT-PCBA data set. OnionNet-SFCT proposed by Zheng et al 33 also incorporated a correction term to AutoDock Vina, but this term was calculated based on a binding pose prediction model rather than binding affinity prediction, implying that the training of the model still relies on the involvement of decoy poses. The linear combination of Vina scores and root-mean-square error (RMSD) values produced by the ML model enabled the method to perform quite well in terms of docking and screening powers.…”

New Wine in an Old Bottle: Utilizing Chemical Genetics to Dissect Apical Hook Development

“…This revolutionary advance not only makes it easier to obtain highly confident predicted protein structure of interest to perform high-throughput virtual screening, but also provides the possibility of reverse docking (RevDock) with a single small-molecule ligand to proteome-wide pockets in one or multiple organisms. With a refined algorithm, a novel protein–ligand scoring function, OnionNet-SFCT, was developed to assist the target search of ligands [ 87 ]. This artificial intelligence (AI) drug discovery and design (AIDD) strategy displayed satisfactory precision, for a well-established plant hormone ABA, which is identified by 14 receptors in planta , 4 of which could be found in the top 10 interacting proteins, and 8 of which could be found in the top 100 proteins [ 87 ].…”

“…With a refined algorithm, a novel protein–ligand scoring function, OnionNet-SFCT, was developed to assist the target search of ligands [ 87 ]. This artificial intelligence (AI) drug discovery and design (AIDD) strategy displayed satisfactory precision, for a well-established plant hormone ABA, which is identified by 14 receptors in planta , 4 of which could be found in the top 10 interacting proteins, and 8 of which could be found in the top 100 proteins [ 87 ]. Taken together, although both CETSA and RevDock possess a certain degree of error rate, we can suppose that the combination of these two methods would be promising in improving the drug target identification.…”

Multitargeted Molecular Docking and Dynamic Simulation Studies of Bioactive Compounds from Rosmarinus officinalis against Alzheimer’s Disease

“…Recent amendments have implemented this functional deficiency in AutoDock Vina 1.2.0 and combined the scoring function of AutoDock 4.2, along with the concurrent docking of multiple ligands and a batch mode for docking a sizeable count of ligands [ 45 ]. Integration of a scoring function correction term improves the protein–ligand docking and screening accuracies that substantially facilitate the prediction abilities for the docking of AutoDock Vina and screening tasks based on CASF-2016, DUD-E and DUD-AD [ 46 ]. Amendments in certain empirical parameters may also improve the ligand ranking of AutoDock Vina [ 47 ].…”