Definitions regarding the base-pair-specific performance of Transat.

In order to quantify the performance of Transat, base-pairs are first classified into true positives (TP), true negatives (TN), false positives (FP) and false negatives (FN) according to the definitions above, where denotes the user-defined p-value threshold which is applied to the helices predicted by Transat. The minimum p-value of a predicted base-pair is defined as the minimum p-value of all predicted helices that contain this base-pair, i.e. a predicted base-pair inherits its statistical significance from the most statistically significant helix to which it belongs.