Fig. 3From: Correlation of gene expression and associated mutation profiles of APOBEC3A, APOBEC3B, REV1, UNG, and FHIT with chemosensitivity of cancer cell lines to drug treatmenta–c Overall motif counts in different cancer types and across all cell lines (pan-cancer analysis). The y axis is presented on the log10 scale. a T(C>K)W motif counts. b T(C>D)R motif counts. c T(C>D)D motif counts. d–f Numbers of distinct, not overlapping 5/1000 kataegis clusters with ≥ 5 motifs on the same genome strand per 1000 bp in different cancer types and in the pan-cancer dataset. d T(C>K)W motif counts. e T(C>D)R motif counts. f T(C>D)D motif counts. Horizontal middle bars show the mean for each cancer category. Vertical bars show mean ± standard deviation. Negative values of (mean − standard deviation) in d and e were truncated at 0. Cancer categories with no vertical columns had no predicted kataegis clusters (d–f) and/or too few cell lines to compute the standard deviation (n = 2 for mesothelioma, a–c)Back to article page