Fig. 1: Mutational signature profile in canine osteosarcoma.

a Distribution of total short variants (SNVs and INDELs) across 26 samples. The samples were sorted in ascending order of disease-free interval (DFI) and binned in three discrete groups. b Distribution of 10 different mutation types as identified by Ensembl Variant Effect Predictor across 26 samples. Missense mutations were the most common type of short variant in this dataset. c Mutation signature analyses showing the relative distribution of six single nucleotide changes in 96 different trinucleotide contexts across three de novo signatures. The de novo signatures were extracted from frequency of single nucleotide changes in 26 tumors using non-negative matrix factorization (NMF) method. d Relative contribution of three predominant de novo signatures across 26 samples. Source data for these plots are included in the Supplementary Data 11.