Fig. 2

Tumor-specific somatic mutations identified from RNA-seq. a Distribution of tumor-specific somatic mutations in different genomic regions. The number of total somatic mutations in various genomic regions and the numbers of recurrent mutations (numbers in yellow stars). Most of the somatic mutations reside in genic regions, and missense mutations in coding sequences accounted for ~ 50% of genic mutations. 5′ Upstream: mutation occurring within 5 kb upstream of genes. 3′ Downsteram: mutation occurring within 5 kb downstream genes. Splice donor: mutation that changes one of 2 bases at the 5′ end of an intron. Splice acceptor: mutation that changes one of 2 bases at the 3′ end of an intron. Splice region: mutation within 1–3 bases of the exon or 3–8 bases of the intron flanking the intron–exon boundary. Deleterious: mutation predicted to be damaging to protein function by both Polyphen-2 and SIFT algorithms. NMD: a mutation predicted to cause nonsense-mediated decay. b Association of mutations in HCC patients with clinical characteristics. The percentage of patients with mutations (Y-axis) for the various genes (X-axis) associated with the various clinical phenotype (X-axis below the genes). Red balls denote bad prognosis (e.g. associated with high Edmondson grade tumor, late stage, necrosis or liver cirrhosis) while green balls represents good prognosis (protective genes associated with no tumor invasion/liver cirrhosis). Size represents significance of association i.e. larger size, smaller p-value. c Association of mutations in HCC patients with tumor size. Box plot show tumor size of patients with and without mutation in WASH1 gene. d Mutations in Genes associated with overall survival. Patients with mutations in KIF21A (left panel) and LSS (right panel) correspond to significantly shorter survival time (p-value < 0.01, Kaplan–Meier test). Green lines represent patients with no mutations in the gene while red lines represent patients with mutations in the gene