Fig. 1

Landscape of long-read transcriptomes in primary and metastatic liver cancers. a. Schematic representation of primary and metastatic liver cancer isoform profiling using Iso-seq and RNA-seq. b. Isoform calling algorithm employed for Iso-Seq data analysis. c. Types of isoforms identified and their graphical illustration. (d-e). The percentage and number of distinct isoforms in each category from (c) are indicated, including total, coding, and noncoding transcripts. f. Characteristics of novel (NIC and NNC) and known (FSM and ISM) transcripts. NIC and NNC isoforms have more exons. g. Proportion of annotated and unannotated junctions based on comparison with reference transcripts. h. Percentage of Iso-seq isoform transcription start sites supported by CAGE (FANTOM5) or transcription termination sites supported by the presence of a poly(A) motif (SQANTI3), plotted per category from (c). i. Global transcript expression patterns of primary and metastatic liver cancers visualized using a t-distributed stochastic neighbor embedding (t-SNE) projection. The position of samples within the plot reflects the relative similarity in transcript expression. Samples are color-coded on the basis of their assigned analysis cohort. T, tumour samples, including primary and metastatic liver cancer samples; Liver, nontumor liver tissues; PT, primary tumour samples of metastatic liver cancer. j. Global transcript expression patterns of liver samples depicted by a t-SNE projection, including HCC-NT, and LM-NT samples