Extended Data Fig. 6: Comparison of 300K-RNA Top-4* with SpliceAI using two additional variant cohorts.
From: SpliceVault predicts the precise nature of variant-associated mis-splicing
a) Mis-splicing events induced by variants in two cohorts curated from literature for additional validation of the Top-4* approach: 58 variants studied in patient specimens and 63 variants studied through midi-gene assays (scrutinized to ensure technical design permitted detection of multi-exon skipping events, see Supplementary Tables 2, 3 for references). b) Sensitivity and PPV of 300K-RNA and SpliceAI for exon-skipping and cryptic activation predictions at different thresholds. Points on the 300K-RNA curve (blue) show metrics when using Top-1*, 2*, 3*, 4* etc events as a prediction of the nature of mis-splicing. Points on the SpliceAI curve (red) show metrics at delta-scores that predict the same number of exon-skipping and cryptic-activation as 300K-RNA top-1*,2*,3*,4* etc to enable direct comparison of true/false positive rates with the Top-4 method. c) The average sensitivity and PPV of 300K-RNA and SpliceAI across three variant cohorts shown in (B). Points on the curves correspond to thresholds used in panel B. * = skipping one or two exons and cryptic activation within 600 nt of the annotated splice-site.
