a, Comparison of short-read and long-read RNA-seq for novel isoform discovery. Short-read RNA-seq relies on spliced junction reads to define splice sites, whereas long-read RNA-seq can sequence an entire isoform in a single read, enabling the precise delineation of isoform structure, identifying novel sequences not present in the reference annotation (red boxes) as well as the potential amino acid coding sequence for each isoform, represented by the wide boxes. Narrow boxes represent 5’ and 3’ untranslated regions. b, Junction usage, as exemplified by Leafcutter, takes a ratio of each junction over the total set of overlapping junctions. c, Isoform usage calculates the ratio of each isoform over the total set of isoforms. d, In splicing events, as exemplified by SUPPA, the isoforms that contain the inclusion of a particular splicing event (in this case, inclusion of novel cassette exon), are summed and divided by the total isoforms for that gene. e, Schematic outline of data generation and analysis in this study. TSS: transcription start site, polyA: polyadenylation site.