Cscore: A Simple Yet Effective Scoring Function for Protein–ligand Binding Affinity Prediction Using Modified Cmac Learning Architecture

Ouyang, Xuchang; Handoko, Stephanus Daniel; Kwoh, Chee Keong

doi:10.1142/s021972001100577x

Cited by 44 publications

(34 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The NNScore series has been largely designed for VS, an application at which these machine‐learning models excel (see Section ‘Machine‐learning SFs for virtual screening’) and thus only provided limited validation for binding affinity prediction. CScore is another NN‐based SF introducing the innovation of generating two features per each atom pair accounting for attraction and repulsion based on a distance‐dependent fuzzy membership function. On PDBbind benchmark, CScore obtained R p = 0.801, a notable improvement over RF‐Score.…”

Section: Generic Machine‐learning Sfs To Predict Binding Affinitymentioning

confidence: 99%

“…Regarding further applications of support vector regression (SVR), Li et al combined SVR with knowledge‐based pairwise potentials as features (SVR‐KB), which outperformed all the classical SFs on the CSAR benchmark by a large margin. An attempt to predict enthalpy and entropy terms in addition to binding energy has also been presented, although the performance of this machine‐learning SF on the PDBbind benchmark is sensibly worse than that of other SVR‐based SFs, suggesting that revising the implementation of this SVR should yield better results on these terms as well. On a posterior study, Ballester introduced SVR‐Score which trained using the same data, features, and protocol as RF‐Score .…”

Section: Generic Machine‐learning Sfs To Predict Binding Affinitymentioning

confidence: 99%

See 1 more Smart Citation

Machine‐learning scoring functions to improve structure‐based binding affinity prediction and virtual screening

Ain

Aleksandrova

Roessler

et al. 2015

WIREs Comput Mol Sci

291

234

View full text Add to dashboard Cite

Docking tools to predict whether and how a small molecule binds to a target can be applied if a structural model of such target is available. The reliability of docking depends, however, on the accuracy of the adopted scoring function (SF). Despite intense research over the years, improving the accuracy of SFs for structure‐based binding affinity prediction or virtual screening has proven to be a challenging task for any class of method. New SFs based on modern machine‐learning regression models, which do not impose a predetermined functional form and thus are able to exploit effectively much larger amounts of experimental data, have recently been introduced. These machine‐learning SFs have been shown to outperform a wide range of classical SFs at both binding affinity prediction and virtual screening. The emerging picture from these studies is that the classical approach of using linear regression with a small number of expert‐selected structural features can be strongly improved by a machine‐learning approach based on nonlinear regression allied with comprehensive data‐driven feature selection. Furthermore, the performance of classical SFs does not grow with larger training datasets and hence this performance gap is expected to widen as more training data becomes available in the future. Other topics covered in this review include predicting the reliability of a SF on a particular target class, generating synthetic data to improve predictive performance and modeling guidelines for SF development. WIREs Comput Mol Sci 2015, 5:405–424. doi: 10.1002/wcms.1225For further resources related to this article, please visit the WIREs website.

show abstract

Section: Generic Machine‐learning Sfs To Predict Binding Affinitymentioning

confidence: 99%

Section: Generic Machine‐learning Sfs To Predict Binding Affinitymentioning

confidence: 99%

Machine‐learning scoring functions to improve structure‐based binding affinity prediction and virtual screening

Ain

Aleksandrova

Roessler

et al. 2015

WIREs Comput Mol Sci

291

234

View full text Add to dashboard Cite

show abstract

“…CScore is an important scoring function for binding affinity prediction [33], which always reports the output of the docking energies as total score. CScore could be converted into binding free energy (ΔG binding = −2.303RT × total score).…”

Section: Resultsmentioning

confidence: 99%

Identification of The Fipronil Resistance Associated Mutations in Nilaparvata lugens GABA Receptors by Molecular Modeling

et al. 2019

View full text Add to dashboard Cite

Fipronil, as the first commercialized member of phenylpyrazole insecticides, has been widely used to control planthoppers in China due to its high insecticidal activity and low toxicity to mammals. However, insects have developed resistance to phenylpyrazoles after their long-term use. The resistance mechanism of insects to fipronil has not been well identified, which limited the development of phenylpyrazole insecticides. In the present study, we aimed to elucidate the related fipronil-resistance mechanism in N. lugens GABA receptors by homology modeling, molecular docking, and molecular dynamics. The results indicated that fipronil showed the weakest interaction with the mutant (R0′Q + A2′S) GABA receptors, which is consistent with the experimental study. The binding poses of fipronil were found to be changed when mutations were conducted. These findings verified the novel fipronil-resistance mechanism in silico and provide important information for the design of novel GABAR-targeting insecticides.

show abstract

“…Scoring functions that exhibit a Pearson correlation >0.72 and an RMSD <2 Å between predicted and experimental binding affinity in cross-validation analyses are commonly characterized as providing robust affinity inferences [ 11 , 40 , 42 – 44 ]. While our results do suggest that incorporating additional structural information can improve protein-protein affinity prediction, the improvements in accuracy we observed were generally incremental, and even best-case accuracy currently remains too low to support robust affinity inferences.…”

Section: Resultsmentioning

confidence: 99%

Improving the accuracy of high-throughput protein-protein affinity prediction may require better training data

Dias

Kolaczkowski

2017

BMC Bioinformatics

View full text Add to dashboard Cite

BackgroundOne goal of structural biology is to understand how a protein’s 3-dimensional conformation determines its capacity to interact with potential ligands. In the case of small chemical ligands, deconstructing a static protein-ligand complex into its constituent atom-atom interactions is typically sufficient to rapidly predict ligand affinity with high accuracy (>70% correlation between predicted and experimentally-determined affinity), a fact that is exploited to support structure-based drug design. We recently found that protein-DNA/RNA affinity can also be predicted with high accuracy using extensions of existing techniques, but protein-protein affinity could not be predicted with >60% correlation, even when the protein-protein complex was available.MethodsX-ray and NMR structures of protein-protein complexes, their associated binding affinities and experimental conditions were obtained from different binding affinity and structural databases. Statistical models were implemented using a generalized linear model framework, including the experimental conditions as new model features. We evaluated the potential for new features to improve affinity prediction models by calculating the Pearson correlation between predicted and experimental binding affinities on the training and test data after model fitting and after cross-validation. Differences in accuracy were assessed using two-sample t test and nonparametric Mann–Whitney U test.ResultsHere we evaluate a range of potential factors that may interfere with accurate protein-protein affinity prediction. We find that X-ray crystal resolution has the strongest single effect on protein-protein affinity prediction. Limiting our analyses to only high-resolution complexes (≤2.5 Å) increased the correlation between predicted and experimental affinity from 54 to 68% (p = 4.32x10−3). In addition, incorporating information on the experimental conditions under which affinities were measured (pH, temperature and binding assay) had significant effects on prediction accuracy. We also highlight a number of potential errors in large structure-affinity databases, which could affect both model training and accuracy assessment.ConclusionsThe results suggest that the accuracy of statistical models for protein-protein affinity prediction may be limited by the information present in databases used to train new models. Improving our capacity to integrate large-scale structural and functional information may be required to substantively advance our understanding of the general principles by which a protein’s structure determines its function.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-017-1533-z) contains supplementary material, which is available to authorized users.

show abstract

Cscore: A Simple Yet Effective Scoring Function for Protein–ligand Binding Affinity Prediction Using Modified Cmac Learning Architecture

Cited by 44 publications

References 24 publications

Machine‐learning scoring functions to improve structure‐based binding affinity prediction and virtual screening

Machine‐learning scoring functions to improve structure‐based binding affinity prediction and virtual screening

Identification of The Fipronil Resistance Associated Mutations in Nilaparvata lugens GABA Receptors by Molecular Modeling

Improving the accuracy of high-throughput protein-protein affinity prediction may require better training data

Contact Info

Product

Resources

About