The impact of cross-docked poses on performance of machine learning classifier for protein–ligand binding pose prediction

Shen, Chao; Hu, Xueping; Gao, Junbo; Zhang, Xujun; Zhong, Haiyang; Wang, Zhe; Xu, Lei; Kang, Yu; Cao, Dong‐Sheng; Hou, Tingjun

doi:10.1186/s13321-021-00560-w

Cited by 30 publications

(39 citation statements)

References 68 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“… a “_Re” represents that the model is trained based on redocked poses in PDBbind-CrossDocked-Refined while “_Cross” represents that the model is trained based on cross-docked poses. Except for DeepDockM and RTMScore, the data of methods is copied from ref . b The top20 success rate of each docking method for a certain data set, and it represents the upper limit of the value in each column. c “re” represents the performance on redocked poses, while “cross” denotes that on cross-docked poses. …”

Section: Resultsmentioning

confidence: 99%

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

Shen

Zhang

Deng³

et al. 2022

J. Med. Chem.

Self Cite

102

View full text Add to dashboard Cite

The past few years have witnessed enormous progress toward applying machine learning approaches to the development of protein–ligand scoring functions. However, the robust performance and wide applicability of scoring functions remain a big challenge for increasing the success rate of docking-based virtual screening. Herein, a novel scoring function named RTMScore was developed by introducing a tailored residue-based graph representation strategy and several graph transformer layers for the learning of protein and ligand representations, followed by a mixture density network to obtain residue–atom distance likelihood potential. Our approach was resolutely validated on the CASF-2016 benchmark, and the results indicate that RTMScore can outperform almost all of the other state-of-the-art methods in terms of both the docking and screening powers. Further evaluation confirms the robustness of our approach that can not only retain its docking power on cross-docked poses but also achieve improved performance as a rescoring tool in larger-scale virtual screening.

show abstract

Section: Resultsmentioning

confidence: 99%

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

Shen

Zhang

Deng³

et al. 2022

J. Med. Chem.

Self Cite

102

View full text Add to dashboard Cite

show abstract

“…Additionally, artificially redocked poses also help models minimize the importance of ligand size. Many attempts have been made to distinguish between crystal and decoy poses, as well as to predict binding affinity with redocked poses 24,[83][84][85][86] . However, considerable care should be taken in loss function and model design because of data imbalance.…”

Section: Conclusion and Discussionmentioning

confidence: 99%

Assessing the generalization abilities of machine-learning scoring functions for structure-based virtual screening

Zhu

Yang

Huang

2022

Preprint

View full text Add to dashboard Cite

In structure-based virtual screening (SBVS), it is critical for machine-learning scoring functions (MLSFs) to capture protein-ligand atomic interaction patterns. We generated a cross-target generalization ability benchmark for protein-ligand binding affinity prediction to assess whether MLSFs could capture these interactions. By focusing on the local domains in protein-ligand binding pockets, we developed standardized pocket Pfam-based clustering (Pfam-cluster) approach for the generalization ability benchmark. Subsequently, 11 typical MLSFs were tested using random cross-validation (Random-CV), protein sequence similarity-based cross-validation (Seq-CV), and pocket Pfam-based cross-validation (Pfam-CV) methods. Surprisingly, all of the tested models showed decreased performance as they were evaluated from Random-CV to Seq-CV to Pfam-CV experiments, without showing satisfactory generalization capacity. Interpretable analysis revealed that predictions on novel targets by MLSFs were relying on buried solvent accessible surface area (SASA)-related features in complex structures. By combining buried SASA-related information with ligand-specific patterns that were only shared among structurally similar compounds, higher performance in Random-CV tests was attained for Random forest (RF)-Score. Based on these findings, we strongly advise assessing the generalization ability of MLSFs with the Pfam-cluster approach and being cautious with the features learned by MLSFs.

show abstract

“…Additionally, artificially redocked poses also help models minimize the dependency on ligand size. Many attempts have been made to distinguish between crystal and decoy poses, and to predict binding affinity with redocked poses 24,[84][85][86][87] . However, caution should be taken in loss function and model design because of data imbalance.…”

Section: Conclusion and Discussionmentioning

confidence: 99%

Assessment of the generalization abilities of machine-learning scoring functions for structure-based virtual screening

Zhu

Yang

Huang

2022

Preprint

View full text Add to dashboard Cite

In structure-based virtual screening (SBVS), it is critical that scoring functions capture protein-ligand atomic interactions. By focusing on the local domains of ligand binding pockets, a standardized pocket Pfam-based clustering (Pfam-cluster) approach was developed to assess the cross-target generalization ability of machine-learning scoring functions (MLSFs). Subsequently, 11 typical MLSFs were evaluated using random cross-validation (Random-CV), protein sequence similarity-based cross-validation (Seq-CV), and pocket Pfam-based cross-validation (Pfam-CV) methods. Surprisingly, all of the tested models showed decreased performances from Random-CV to Seq-CV to Pfam-CV experiments, not showing satisfactory generalization capacity. Our interpretable analysis suggested that the predictions on novel targets by MLSFs were dependent on buried solvent-accessible surface area (SASA)-related features of complex structures, with larger predicted binding affinities on complexes owning larger protein-ligand interfaces. By combining buried SASA-related features with complex-specific patterns that were only shared among structurally similar compounds in the same cluster, random forest (RF)-Score attained a good performance in Random-CV test. Based on these findings, we strongly advise to assess the generalization ability of MLSFs with Pfam-cluster approach and to be cautious with the features learned by MLSFs.

show abstract

The impact of cross-docked poses on performance of machine learning classifier for protein–ligand binding pose prediction

Cited by 30 publications

References 68 publications

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

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

Assessing the generalization abilities of machine-learning scoring functions for structure-based virtual screening

Assessment of the generalization abilities of machine-learning scoring functions for structure-based virtual screening

Contact Info

Product

Resources

About