Abstract:Structure-based drug design depends on the detailed knowledge of the three-dimensional (3D) structures of protein–ligand binding complexes, but accurate prediction of ligand-binding poses is still a major challenge for molecular docking due to deficiency of scoring functions (SFs) and ignorance of protein flexibility upon ligand binding. In this study, based on a cross-docking dataset dedicatedly constructed from the PDBbind database, we developed several XGBoost-trained classifiers to discriminate the near-native binding poses from decoys, and systematically assessed their performance with/without the involvement of the cross-docked poses in the training/test sets. The calculation results illustrate that using Extended Connectivity Interaction Features (ECIF), Vina energy terms and docking pose ranks as the features can achieve the best performance, according to the validation through the random splitting or refined-core splitting and the testing on the re-docked or cross-docked poses. Besides, it is found that, despite the significant decrease of the performance for the threefold clustered cross-validation, the inclusion of the Vina energy terms can effectively ensure the lower limit of the performance of the models and thus improve their generalization capability. Furthermore, our calculation results also highlight the importance of the incorporation of the cross-docked poses into the training of the SFs with wide application domain and high robustness for binding pose prediction. The source code and the newly-developed cross-docking datasets can be freely available at https://github.com/sc8668/ml_pose_prediction and https://zenodo.org/record/5525936 , respectively, under an open-source license. We believe that our study may provide valuable guidance for the development and assessment of new machine learning-based SFs (MLSFs) for the predictions of protein–ligand binding poses.

zPoseScore model for accurate and robust protein–ligand docking pose scoring in CASP15

Estimating protein complex model accuracy based on ultrafast shape recognition and deep learning in CASP15.

Template‐guided method for protein–ligand complex structure prediction: Application to CASP15 protein–ligand studies

DeepBSP—a Machine Learning Method for Accurate Prediction of Protein–Ligand Docking Structures

GSScore: a novel Graphormer-based shell-like scoring method for protein–ligand docking

A Generalized Protein-Ligand Scoring Framework with Balanced Scoring, Docking, Ranking and Screening Powers.

Generic protein–ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling

A fully differentiable ligand pose optimization framework guided by deep learning and a traditional scoring function

Accurate Protein-Ligand Complex Structure Prediction using Geometric Deep Learning

Protein inter-residue contact and distance prediction by coupling complementary coevolution features with deep residual networks in CASP14

Integrating deep learning, threading alignments, and a multi‐MSA strategy for high‐quality protein monomer and complex structure prediction in CASP15

Deep Scoring Neural Network Replacing the Scoring Function Components to Improve the Performance of Structure-Based Molecular Docking.

The Impact of Cross-Docked Poses on Performance of Machine Learning Classifier for Protein–ligand Binding Pose Prediction

Synergistic Application of Molecular Docking and Machine Learning for Improved Binding Pose

Conformation search across multiple-level potential-energy- surfaces (CSAMP): A strategy for accurate prediction of protein-ligand binding structures.

A fully differentiable ligand pose optimization framework guided by deep learning and traditional scoring functions

High‐accuracy protein structure prediction in CASP14

Improving Protein-Peptide Docking Results Via Pose-Clustering and Rescoring with a Combined Knowledge-Based and MM-GBSA Scoring Function.

EquiScore: A generic protein-ligand interaction scoring method integrating physical prior knowledge with data augmentation modeling

Accurate prediction of protein–ligand interactions by combining physical energy functions and graph-neural networks

Accurate prediction of dynamic protein–ligand binding using P‐score ranking