Comparison of Machine Learning Models for the Androgen Receptor

Zorn, Kimberley M.; Foil, Daniel H.; Lane, Thomas R.; Hillwalker, Wendy; Feifarek, David J; Jones, Frank E.; Klaren, William D; Brinkman, Ashley M; Ekins, Sean

doi:10.1021/acs.est.0c03984

Cited by 21 publications

(20 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…29 This study may be enlightening as it echoes our earlier findings on individual datasets after extensive manual curation. 52,53,[55][56][57][58][59][60]82 It also goes some way further in using external validation with these methods for toxicology and drug discovery properties. Now that we have generated such a vast array of over 5000 machine learning models, it presents opportunities for using them for predicting the potential of new molecules to interact with targets of interest or avoiding others (such as PXR and hERG) that may result in undesirable effects.…”

Section: ■ Discussionmentioning

confidence: 99%

“…All of the preceding examples have in common that few, if any, used prospective prediction as a form of validation. The main aim of this study was to evaluate a Bayesian method, Assay Central, which we have previously used in several examples where we have compared it versus other machine learning methods against a relatively small number of datasets or targets such as drug-induced liver injury, rat acute oral toxicity, estrogen receptor, , androgen receptor, GSK3β, Mycobacterium tuberculosis, , non-nucleoside reverse transcriptase, and whole cell HIV . Our evaluation has now been expanded with this study to over 5000 CHEMBL datasets and provides statistics and data visualization methods in order to assess each algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Bioactivity Comparison across Multiple Machine Learning Algorithms Using over 5000 Datasets for Drug Discovery

et al. 2020

Self Cite

View full text Add to dashboard Cite

Machine learning methods are attracting considerable attention from the pharmaceutical industry for use in drug discovery and applications beyond. In recent studies, we and others have applied multiple machine learning algorithms and modeling metrics and, in some cases, compared molecular descriptors to build models for individual targets or properties on a relatively small scale. Several research groups have used large numbers of datasets from public databases such as ChEMBL in order to evaluate machine learning methods of interest to them. The largest of these types of studies used on the order of 1400 datasets. We have now extracted well over 5000 datasets from CHEMBL for use with the ECFP6 fingerprint and in comparison of our proprietary software Assay Central with random forest, k-nearest neighbors, support vector classification, naïve Bayesian, AdaBoosted decision trees, and deep neural networks (three layers). Model performance was assessed using an array of fivefold cross-validation metrics including area-under-the-curve, F1 score, Cohen’s kappa, and Matthews correlation coefficient. Based on ranked normalized scores for the metrics or datasets, all methods appeared comparable, while the distance from the top indicated that Assay Central and support vector classification were comparable. Unlike prior studies which have placed considerable emphasis on deep neural networks (deep learning), no advantage was seen in this case. If anything, Assay Central may have been at a slight advantage as the activity cutoff for each of the over 5000 datasets representing over 570,000 unique compounds was based on Assay Central performance, although support vector classification seems to be a strong competitor. We also applied Assay Central to perform prospective predictions for the toxicity targets PXR and hERG to further validate these models. This work appears to be the largest scale comparison of these machine learning algorithms to date. Future studies will likely evaluate additional databases, descriptors, and machine learning algorithms and further refine the methods for evaluating and comparing such models.

show abstract

Section: ■ Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Bioactivity Comparison across Multiple Machine Learning Algorithms Using over 5000 Datasets for Drug Discovery

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…(e) (f) Using 20 gold-standard reference androgen receptor probe compounds as used by Kleinstreuer et al [15] shows that there was a good result for predictions of 16/20, i.e., 80% were predicted correct for being a binder to AR (very weak binders were considered as non-binders). With respect to well-known compounds that are frequently misclassified [16], the results provided here (Table 3) show four out of 11 compounds correctly predicted compared to three out of 11 reported elsewhere, the difference being the correct prediction of finasteride [16]. Chemical structures in Tables 2 and 3 show several steroid core structures that may be difficult for algorithms to distinguish between actives and inactives, given the strong chemical similarity between them.…”

Section: Resultsmentioning

confidence: 60%

Androgen Receptor Binding Prediction with Random Forest, Deep Neural Networks, and Graph Convolutional Neural Networks

García-Sosa¹

2021

Preprint

View full text Add to dashboard Cite

Substances that can modify the androgen receptor pathway in humans and animals are entering the environment and food chain with the proven ability to disrupt hormonal systems and leading to toxicity and adverse effects on reproduction, brain development, and prostate cancer, among others. State-of-the-art databases with experimental data of human, chimp, and rat effects by chemicals have been used to build machine learning classifiers and regressors and evaluate these on independent sets. Different featurizations, algorithms, and protein structures lead to different results, with deep neural networks on user-defined physicochemically-relevant features developed for this work outperform graph convolutional, random forest, and large featurizations. The results can help provide clues on risk of substances and better experimental design for toxicity assays. Source code and data are available at https://github.com/AlfonsoTGarcia-Sosa/ML

show abstract

“…DNNs are not always superior to other methods in predictive performance, as shown in previous studies. 57 , 58 DNNs are susceptive of hyperparameters, architecture, and optimizers. It is difficult to achieve competitive performance with other methods in certain cases.…”

Section: Resultsmentioning

confidence: 99%

Direct Prediction of Physicochemical Properties and Toxicities of Chemicals from Analytical Descriptors by GC–MS

Zushi

2022

Anal. Chem.

View full text Add to dashboard Cite

With advances in machine learning (ML) techniques, the quantitative structure–activity relationship (QSAR) approach is becoming popular for evaluating chemicals. However, the QSAR approach requires that the chemical structure of the target compound is known and that it should be convertible to molecular descriptors. These requirements lead to limitations in predicting the properties and toxicities of chemicals distributed in the environment as in the PubChem database; the structural information on only 14% of compounds is available. This study proposes a new ML-based QSAR approach that can predict the properties and toxicities of compounds using analytical descriptors of mass spectrum and retention index obtained via gas chromatography–mass spectrometry without requiring exact structural information. The model was developed based on the XGBoost ML method. The root-mean-square errors ( RMSE s) for log K o-w , log (molecular weight), melting point, boiling point, log (vapor pressure), log (water solubility), log (LD 50 ) (rat, oral), and log (LD 50 ) (mouse, oral) are 0.97, 0.052, 51, 23, 0.74, 1.1, 0.74, and 0.6, respectively. The model performed well on a chemical standard mixture measurement, with similar results to those of model validation. It also performed well on a measurement of contaminated oil with spectral deconvolution. These results indicate that the model is suitable for investigating unknown-structured chemicals detected in measurements. Any online user can execute the model through a web application named Detective-QSAR ( ). The analytical descriptor-based approach is expected to create new opportunities for the evaluation of unknown chemicals around us.

show abstract

Comparison of Machine Learning Models for the Androgen Receptor

Cited by 21 publications

References 51 publications

Bioactivity Comparison across Multiple Machine Learning Algorithms Using over 5000 Datasets for Drug Discovery

Bioactivity Comparison across Multiple Machine Learning Algorithms Using over 5000 Datasets for Drug Discovery

Androgen Receptor Binding Prediction with Random Forest, Deep Neural Networks, and Graph Convolutional Neural Networks

Direct Prediction of Physicochemical Properties and Toxicities of Chemicals from Analytical Descriptors by GC–MS

Contact Info

Product

Resources

About