Fig. 3: Mutational signature clustering and analysis reaffirms the importance of APOBEC in UC.

a Heatmap of mutational signatures with cosine similarity (CS) > 0.25 in at least 10% of samples. The samples (n = 191 samples) were sorted by the consensus cluster plus (CCP) clusters and clustered by signature. b, c Boxplots were generated for the CS of SBS13 (b) and CS of SBS2 (c) for patients divided into TMB tertiles (n = 191 samples). d Boxplot of TMB by K1 and K2 CCP cluster (n = 191 samples). e SBS2 and f SBS44 were plotted by variant histology for those samples with annotated histology (n = 147 samples). All boxplots are shown with boxes representing the IQR and midline at the median. Error bars represent Q1/Q3 ± 1.5*IQR. Two-sided Wilcoxon test p-values are shown above the given comparison. Kaplan-Meier curves were used to visualize survival from time of treatment initiation for g chemotherapy and h immunotherapy. Samples were split into high, medium, and low APOBEC activity based on the tertiles of the rank order CS for SBS13. Cox proportional-hazard modeling was performed with the high group as reference, Hazard Ratio (95% CI) and adjusted p-value for each comparison are inset with risk tables below. Source data are provided as a Source Data file.