Figure 2

SVM classification of Mouse/Human CSTS in training and validation. A: ROC curves for long CST training (Yellow), long CST validation (Red), short CST training (Green) and short CST validation (Blue) experiments. B: P-score distributions of Coding and Non coding CSTs. The x axis represents the P-score generated by the SVM, the y axis shows relative frequency of CSTs by dataset (short coding, short non-coding, long coding and long non-coding) for training and validation phases of the experiment colour keys as in A, thick lines correspond to non-coding CSTs, fine lines correspond to coding CSTs.