Project: PRJEB61296
In this study, we employed a combination of short-read and PacBio Iso-Seq-based long-read scRNA-seq on colorectal cancer clinical samples to construct a comprehensive transcriptomic atlas, enabling the exploration of isoform-specific dysregulations in CRC. With this atlas, (1) we identified 394 DTSs across 273 genes in tumor epithelial cells, with a significant number arising from combinations of multiple AS events, which could only be achieved with LR-seq data and thus have been poorly investigated; (2) we unveiled genes and transcripts linked to the three normal colon epithelial lineages and three tumor epithelial subpopulations that are correlated with unique patient prognoses; (3) by incorporating the novel peptides derived from predicted ORFs of recurrent tumor-specific transcripts and mass spectrometry data, we devised an algorithm to generate a set of 24 neoantigens exhibiting high binding affinities to major histocompatibility complex (MHC) molecules for a diverse group of CRC patients.
General