Fig. 2

Characteristics of long-read ONT novel transcripts. A Percentage of long-read ONT transcripts with TSS 100 bp away from known TSS (left), with TTS 100 bp away from known TTS (center), and with predicted nonsense-mediated mRNA decay (right). B Correlation of transcript-level expression calculated by short-read and long-read sequencing. C The average expression distribution of different types of transcripts, Wilcoxon signed-rank test, *** P < 0.001. D Pathways significantly enriched for genes with an elevated number of novel transcripts detected by long-read ONT sequencing in CRC. Bubble size denotes the number of genes with novel transcripts in each pathway, and color denotes the significance. E Enrichment analysis of oncogenes and tumor suppressor genes with novel transcripts detected by long-read ONT sequencing in cancer. F Number of novel long-read ONT transcripts compared to annotated GENCODE v34 transcripts for selected oncogenes. G Correlation between phyloP score and phastCon score of novel transcripts. H Percent of long-read isoforms predicted to contain an ORF by Transdecoder. I Number of predicted ORFs validated by CPTAC LC–MS/MS proteomics data