Fig. 4: Performance evaluation of SpliceAI and Pangolin using merged Var.GWAS and Var.P datasets.
From: Functional impact of splicing variants in the elaboration of complex traits in cattle
a Distribution of spliceAI and Pangolin absolute scores in Var.GWAS and Var.P datasets. b Scatter plot of SpliceAI versus Pangolin scores. SpliceAI scores related to donor or acceptor loss have been turned into opposite numbers to allow comparison with Pangolin scores. r Pearson correlation coefficient. c Scatter plot of SpliceAI and Pangolin scores each versus ∆PSI values in both HEK293T or MAC-T cell lines. d True positive and true negative rates of SpliceAI and Pangolin obtained and classified by absolute score category. SDV identified using Vex-seq were used to calculate the true positive rate, and non-SDV to calculate the true negative rate. TP true positive, TN true negative, SpAI SpliceAI, Pan Pangolin. e Receiving operating curve (ROC) and area under the curve (AUC) were calculated for SpliceAI and Pangolin in each cell line and for the mean values between the two cell lines. Source data are provided as a Source Data file.
