Fig. 2From: Predicting variant deleteriousness in non-human species: applying the CADD approach in mousea-d) ROC-AUC scores of the four different hCADD models evaluated on the human held-out test set. e, f) Seven different subsets of the human held-out test set evaluated by primate- and vertebrate based PhastCons scores, respectively. g) Missense mutations of the human held-out test set evaluated by SIFT. (see caption Fig. 1 for remaining explanation)Back to article page