eNetXplorer: an R package for the quantitative exploration of elastic net families for generalized linear models

SummarySystems biology analysis often involves building predictive models by selecting informative features from a large number of measurements. The elastic net for generalized linear models is a popular regression and feature selection method, particularly useful when the number of features is greater than the sample size or when there exist many correlated predictor variables. This package provides a quantitative, cross-validation based toolkit to evaluate elastic net models and to uncover correlates contributing to prediction. Feature importance is evaluated by flexible criteria using out-of-bag prediction performance assessed via user-defined quality functions. Statistical significance is assigned to each model by comparison to null models generated by permutations of sample labels; analogous approaches are used to assess significance for the contribution of individual features to prediction. This package fits linear, binomial (logistic) and multinomial models, and provides a set of standard plots, summary statistics and output tables. eNetXplorer enables quantitative, exploratory analysis to generate hypotheses on which features may be associated with biological phenotypes of interest, such as in the identification of biomarkers for therapeutic responsiveness.\n\nAvailability and implementationThe eNetXplorer R package is available under GPLv3 license at https://CRAN.R-project.org/package=eNetXplorer