Fig. 6

Long-read-informed protein isoform detection. a Scenarios in which long-read transcriptomic abundance values can be used to “rescue and resolve” protein group identifications for improved protein isoform detection. Note that peptide locations are not drawn to scale. See Ref. [10] for original cases. b The number of protein groups containing PacBio-derived protein isoforms identified in a traditional (normal) search versus one that incorporates long-read RNA-seq data to rescue isoforms (Rescue & Resolve algorithm). c Example of two rescued isoforms from the gene IFI16. PB.1137.5 and PB.1137.24 both have high transcriptional abundance and contain novel exon skipping events, but both lack unique sequence regions containing an identified peptide. CPM, full-length copies per million reads. d, e The percent of rescued (d) and resolved or removed (e) protein isoforms validated compared to the derived null distribution from randomly rescued isoforms