Random forest classifier achieves 94% accuracy on mouse tRNA genes. Receiver operating characteristic curves for random forest (blue), logistic regression (red), and support vector machine (yellow) upon application to human training data with 10-fold cross-validation (A) and mouse test data (B) are shown. The number of mouse tRNA genes predicted as active (C) and inactive (D) are compared to the number of tissues in which they are actively transcribed according to Bogu et al. 2016. We considered a mouse tRNA gene active if it is actively transcribed in at least one tissue.