Fig. 1

Whole-exome sequencing in ANKL. a Percentages of somatic base substitutions and indel mutations identified by whole-exome sequencing in tumor-normal paired samples of ANKL (n = 4), NKTCL (n = 25), CLPD-NK (n = 3), T-LGLL (n = 15) and T-PLL (n = 4). Synonymous mutations were included in the analysis. b Normalized weights of trinucleotide signatures identified using deconstructSigs in tumor-normal paired samples of ANKL, NKTCL, CLPD-NK, T-LGLL, and T-PLL. Weights of three most frequent signatures in each cancer type are shown across cancers as separate signatures and others are included under “other”. Synonymous mutations were included in the analysis. c Numbers of somatic mutations in tumor-normal paired samples of ANKL, NKTCL, CLPD-NK, T-LGLL, and T-PLL. Synonymous mutations were included in the analysis. Horizontal lines indicate median, error bars indicate 10th and 90th percentiles, boxes represent interquartile ranges, and dots indicate outliers. P values were calculated using the Mann–Whitney U-test. d Alterations identified by whole-exome sequencing selected based on recurrence and biological significance. Complete lists of identified mutations are found in Supplementary Data 2. Diagonally split dual-colored rectangle indicates the presence of two alterations of different type in the same sample. Reads mapping to the EBV genome are reported as counts per million (CPM) under the figure. Results of the MutSigCV and Oncodrive-fm driver gene analyses are presented on the right side of the figure. Expression estimates of mutated genes in normal NK cells and NK cell lines are shown on the right as reads per kilobase per million mapped reads (RPKM), with bar length indicating mean and error bars representing range