To evaluate how much the genetic variants could explain the splicing outcome for each SE, we performed leave-one-out cross validation. The SNV-determined proportion in the PSI of an exon skipping event was assessed using Pearson’s correlation r between the predicted PSI (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{{\hat{\mathrm \Psi }}}}$$\end{document}Ψ^) and RNA-seq measured PSI (Ψ). An SE with a significant p value for a positive r indicates that the splicing outcome can be, at least partially, explained by the SNVs in the transcribed region. The variability of the PSI explained by the model, R2, was also calculated (Supplementary Fig. S1). Figure 1B shows the quantile-quantile plot of the observed p values from our models (cross validated) against the expected p values under the null hypothesis, which were randomly drawn from a uniform distribution ranging from 0 to 1. We observed a substantial deviation from the null distribution, indicating that exon inclusion of a large proportion of skipped exon events in the DLPFC transcriptome can be partially explained by the genetic variants. We used a Bonferroni p cutoff
