Fig. 4

Detection of annotated and cryptic alternative splicing events across ALS patient brain. (a) MAJIQ was used to detect transcriptome-wide alternative splicing alterations in ALS patients relative to controls (significance threshold of (ΔΨ > 10%, Wilcoxon p-value < 0.05). Here, ALS-implicated genes that had significant alternative splicing events are highlighted per brain region. Asterix indicates that de novo junctions were involved in alternative splicing. ALS-implicated genes with no splicing were ANG, CCNF, CHCHD10, CHMP2B, HNRNPA1, NEFH, OPTN, PFN1, SETX, SPG11, TARDBP, TUBA4A and UBQLN2. (b) VOILA modulizer categorisation of significant alternative splicing events. See Vaquero-Garcia et al. [59]. for definition of event types. “p_” indicates putative. (c) Venn diagram showing the detection of literature-reported cryptic splicing events in each brain region. A cryptic splicing event was considered detected if identified by MAJIQ (ΔΨ > 1%, Wilcoxon p-value < 0.05, de_novo_junction = 1) or LeafCutter (ΔΨ > 1%, FDR < 0.05, classified as unannotated). Visual representation of the (d)RNASET2 and (e)STMN2 cryptic splicing events. The adjacent violin plot indicates the inclusion level of the of the splice junction [E(Ψ)] in ALS motor cortex (MC_ALS) versus control motor cortex (MC_CTRL), relative to other splice junctions in the same local splicing variation (LSV). MC, motor cortex; FC, prefrontal cortex; HP, hippocampus; OC, occipital cortex; CB, cerebellum