Extended Data Fig. 1: Clonality, mutation burden extended and telomere lengths.

a, SNV burden, adjusted for sensitivity using sequencing depth and clonality, against the median VAF in microdissections from non-cancer donors, showing no residual correlation between clonality and SNV burden. P-value obtained from Pearson correlation test. b, median VAF in microdissections from all donors, annotation for the presence of intestinal metaplasia, showing that metasplastic glands have significantly higher median VAFs. P-value obtained through a two-sided Wilcoxon rank sum test. The central line, box and whiskers represent the median, IQR from first to third quartiles, and 1.5 × IQR, respectively. c, SNV mutation burden in glands with confirmed H. pylori status. d-e, Detected burden of (d) SNVs and (e) indels in microdissections from gastric cancers of PD41759 and PD41762, with the clonal SNV and indel burden, present in all cancer samples, denoted by the black dot. The red dashed line indicates the estimated age and SNV and indel mutation burden relation estimated from a mixed effects model in gastric glands from non-cancer donors (Fig. d), with the grey area indicating a confidence interval. (f) Estimated mean telomere length versus age for gastric glands. The red dashed line indicates the maximum likelihood age and telomere length relation estimated from a mixed effects model, with the grey shaded area indicating the 95% confidence interval. Glands with moderate or severe chronic inflammation show significantly shorter telomeres than those without. P-values obtained through a two-sided ANOVA test.