Extended Data Fig. 6: Driver mutations in PNE samples.
From: Age-related remodelling of oesophageal epithelia by mutated cancer drivers

a, Distribution of mutations in 157 PNE and 519 ESCC samples is shown for driver genes significantly mutated in PNE or ESCC samples, including NOTCH2, NOTCH3, CREBBP, FAT1, CHEK2, PAX9, EP300 and PIK3CA. Mutations are depicted separately for PNE (top) and ESCC (bottom). b, Frequencies of driver genes in paired cancer and PNE samples from 68 patients with ESCC. Genes in which mutations were observed in 5% or more samples either in PNE or ESCC were evaluated. Significantly differentially mutated genes (q < 0.05) between PNE and ESCC samples are indicated by asterisks. c, d, Number of driver mutations (c) and their maximum MCFs (d) in samples from high-risk and low-risk individuals are plotted against the age of subject, according to sample size. Regression lines are provided with R2. P values for the significance of lifestyle ESCC risks are also indicated (one-sided Mann–Whitney U-test). e, h, Mean (± 95% confidence interval) of standardized difference of mutation number (e) and maximum MCF (h) between samples from high-risk and low-risk individuals are plotted for each sampling size, and combined samples, using a random-effects model. P value in the random-effects model is also indicated, together with the weight from each sample size (in per cent) for the model fitting (two-sided Wald test) (Methods). f, g, Mean (± 95% confidence interval) of standardized residuals of the number of indicated driver mutations (f) and their maximum MCFs (g) in samples from high-risk individuals against the linear regression model in samples from low-risk individuals (Methods); plotted for 0.2-mm2, 0.8-mm2 and 4-mm2 samples. Standardized residuals for combined samples using a random-effects model are also shown. P-value in the random-effects model is also indicated, together with the weight from each sample size (in per cent) for the model fitting (two-sided Wald test) (Methods). In the analyses in c–h, the numbers of samples from low-risk and high-risk individuals, respectively, are 34 and 19 (0.2-mm2 sampling), 19 and 12 (0.8-mm2 sampling), and 40 and 33 (4-mm2 sampling) (c, e, f); 26 and 18 (0.2-mm2 sampling), 13 and 12 (0.8-mm2 sampling), and 30 and 31 (4-mm2 sampling) (d, h); 5 and 5 (0.2-mm2 sampling), 4 and 7 (0.8-mm2 sampling), and 8 and 19 (4-mm2 sampling) (TP53 in g); 14 and 18 (0.2-mm2 sampling), 7 and 11 (0.8-mm2 sampling), 21 and 28 (4-mm2 sampling) (NOTCH1 in g); 2 and 4 (0.2-mm2 sampling), 2 and 9 (4-mm2 sampling) (PPM1D in g).