Fig. 7: GIE association with cancer genomic features.

a, Heatmap displaying the association of 40 genomic features with GIE frequency across 26 cancer types. The features displayed have, at least, one significant cancer-type association with GIE alterations. Significant associations that cannot be explained by higher background mutation rate are highlighted by a red border. Dot colors are colored according to the log₂(OR). UV, ultraviolet. b, Left to right, dot plot representations of the TMB, clonal TMB and subclonal TMB log₂(OR) across the 26 cancer types. The black square represents the mean values and the error bar the 95% and 5% CIs across the 26 cancer types. The horizontal lines represent a neutral scenario with log₂(OR) = 0 (n is the number of cancer types). c, Analogous representation for predicted neoepitopes, clonal neoepitopes and subclonal neoepitopes. d, Comparison of the APOBEC mutational exposure between samples bearing GIE alterations (GIE) and wild-type (no GIE) in breast cancer. e,f, Analogous comparison for UV-induced double-base substitutions (DBSs) in skin melanoma (e) and for platinum treatment-attributed DBSs in NSCLC (f). g, Comparison of immune infiltration estimates from Davolit et al.³⁷ between samples bearing GIE alterations (GIE) and wild-type (no GIE) in colorectal cancer. Boxplots: the center line is the median and the first section out from the center line contains 50% of the data. The next sections contain half the remaining data until we are at the outlier level. Each level out is shaded lighter. P values of the boxplots are calculated using a two-sided Mann–Whitney U-test. SBS, single base substitution; Suspect., suspected.