Fig. 4

Composite figure demonstrating various analytical aspects of the study. a Venn diagram showcasing overlapping differentially expressed alternative splicing (DEAS) events among the three datasets. b Sashimi plot of the overlapping ZWINT retained intron (RI) event, derived from the PRJEB2449 dataset. The x-axis indicates genomics locations, while the y-axis indicates normalised fragments per kilobase of transcript per million mapped reads (FPKM) values, averaged across samples within each group. The bulk/ ‘sashimi-like’ region indicates a heavily transcribed, i.e. exonic, region. The gaps between these exonic regions indicate the presence of intronic regions. Red and blue sections symbolise grouped tumour and normal samples, respectively. Junction reads are shown as curved lines crossing the exons, with their numbers indicated on the corresponding curves. The averaged percent-spliced-in (PSI) value, calculated within each sample group, is shown on the right as ‘IncLevel’. The bottom black panel represents the alternative exon–intron structures. c Box plot shows the PSI values of the ZWINT RI event in the PRJEB2449 dataset, comparing tumour samples with their matched normal counterparts. Boxplots d and e display the PSI values of the ZWINT RI event in the TCGA-PRAD set, comparing tumour samples versus normal ones, and in the Clariom D set, comparing malignant samples versus matched benign ones, respectively. The unadjusted p or the Benjamini–Hochberg (BH) false discovery rate (FDR) values on the box plots were derived from the results of the corresponding differential splicing analyses conducted using the respective tools. f Bubble plot of the Gene Ontology Biological Process (GO BP) functional enrichment analysis performed on the parent genes of the 141 DEAS events that overlap between the PRJEB2449 and TCGA-PRAD cohorts. g Gene-concept network presents the top five significant terms and the associated genes. TCGA The Cancer Genome Atlas, PRAD prostate adenocarcinoma