Abstract
Support vector machines are a popular machine learning method for many classification tasks in biology and chemistry. In addition, the support vector regression (SVR) variant is widely used for numerical property predictions. In chemoinformatics and pharmaceutical research, SVR has become the probably most popular approach for modeling of non-linear structure-activity relationships (SARs) and predicting compound potency values. Herein, we have systematically generated and analyzed SVR prediction models for a variety of compound data sets with different SAR characteristics. Although these SVR models were accurate on the basis of global prediction statistics and not prone to overfitting, they were found to consistently mispredict highly potent compounds. Hence, in regions of local SAR discontinuity, SVR prediction models displayed clear limitations. Compared to observed activity landscapes of compound data sets, landscapes generated on the basis of SVR potency predictions were partly flattened and activity cliff information was lost. Taken together, these findings have implications for practical SVR applications. In particular, prospective SVR-based potency predictions should be considered with caution because artificially low predictions are very likely for highly potent candidate compounds, the most important prediction targets.
Highlights
Support vector machines (SVMs) are algorithms for supervised machine learning [1] that have become increasingly popular for object classification and ranking in bioinformatics [2,3] and chemoinformatics [4,5], given their often observed high predictive performance compared to other machine learning approaches [5]
The effects of regularization term variations on support vector regression (SVR) model performance were evaluated in detail
We have analyzed in detail the use of SVR models for compound potency prediction, which represents an increasingly popular quantitative structure-activity relationship analysis (QSAR) strategy
Summary
Support vector machines (SVMs) are algorithms for supervised machine learning [1] that have become increasingly popular for object classification and ranking in bioinformatics [2,3] and chemoinformatics [4,5], given their often observed high predictive performance compared to other machine learning approaches [5]. SVMs are often used in combination with kernel functions, which project training sets into feature spaces of higher dimensionality where a linear separation of PLOS ONE | DOI:10.1371/journal.pone.0119301. Support Vector Regression-Based Compound Potency Prediction positive and negative training data might be feasible. In addition to classification and ranking, the SVM approach has been adapted for prediction of numerical property values through support vector regression (SVR) [6,7]. Instead of constructing a hyperplane for classification, SVR derives a function on the basis of training data to predict numerical values. SVR is an intrinsically non-linear prediction approach because it projects data sets characterized by the presence of non-linear structure-property relationships in original feature spaces into higher-dimensional space representations where a linear regression function can be fitted. QSAR has been, and continues to be, the most widely applied computational approach for potency prediction and compound design in medicinal chemistry
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
