Quality control of samples

Potential DNA cross-contamination was checked using mtDNA variant calls. Two samples carrying more than ten heteroplasmic variants with similar HFs were excluded from this study. We also removed four sequences where the average depth of mtDNA was below 400×. A further four iPSCs were removed because their matched fibroblasts were unavailable or failed quality controls. After all of the sample QC steps, 146 fibroblasts and 141 iPSCs were included in the analysis. From 25 donors one iPSC line was included and from 58 donors 2 iPSC lines were included (Fig. 1a).

Although the average depth of mtDNA sequencing was clustered into two groups (Supplementary Fig. ^2j), there was no detectable difference in the depth between paired fibroblasts and iPSCs (median depth in fibroblasts 777×, median depth in iPSCs 887×, P = 0.907, paired Wilcoxon test). Likewise, the average number of heteroplasmic variants, the heteroplasmy fraction (HF), and the average number of low-level heteroplasmic variants (HF < 5% or HF < 10%) were no detectable difference between the two clusters (P = 0.07, P = 0.19, P = 0.10 and P = 0.10, two-sided Wilcoxon rank-sum test) (Supplementary Fig. ^{2g, h}). There was no detectable correlation between either the mean number of heteroplasmic variants or the average heteroplasmy fraction (HF) and the mtDNA sequencing depth (Supplementary Fig. ^{2k, l}).