Extended Data Fig. 3: Within-clone analysis of mtDNA-nuDNA ratio in response to whole-genome doubling.

a, Total number and proportion of diploid and tetraploid cells plotted for each sample. b, Comparison of mtDNA copy number between diploid and tetraploid cells across all 9 184h-TERT cell lines and 7 tumor samples as well as GM18507 lymphoblastoid cells (two-sided, two-sample Wilcoxon test, all P < 7.7 × 10⁻⁹). All boxplots represent the median, 25th percentile and 75th percentile, and whiskers correspond to 1.5 times the interquartile range. c, Violin plot of MNR in diploid (n = 3,475) and tetraploid (n = 350) cells across all four 184-hTERT breast epithelial cell lines. There is no significant difference between the two groups (two-sided, two-sample Wilcoxon test, P = 0.067). All boxplots represent the median, 25th percentile and 75th percentile, and whiskers correspond to 1.5 times the interquartile range. d, Boxplot of the mtDNA copy number distribution of cell cycle-sorted cells in different phases, G1, S and G2, across two sequencing libraries of T-47D breast cancer cell line, SA1044-A96139A (n = 735 cells) and SA1044-A96147A (n = 823 cells) and of lymphoblastoid cell line, SA928-73044A (n = 481 cells) and SA928-A90553C (n = 1,016 cells). All boxplots represent the median, 25th percentile and 75th percentile, and whiskers correspond to 1.5 times the interquartile range. Pairwise significance is indicated by two-sided Wilcoxon tests. e, Same as d, but for MNR. f, Bar plot of the median diameter, measured in microns, for both diploid and tetraploid cells for each library across the 12 sequencing libraries of tumor samples. g, Boxplot of the median, 25th percentile and 75th percentile of fold change of median diameter for tetraploid cells over diploid cells across the same 12 sequencing libraries in f. The whiskers correspond to 1.5 times the interquartile range. h, Graphical model of MityBayes. MityBayes takes raw counts of the alternate allele and total depth per cell across mtDNA variants. It infers the clonal assignment of the cells based on clone-specific heteroplasmy level and weighing of informative variants. i, Heatmap indicating the presence of mtDNA variant, m.1429C>T. Each cell indicates a fraction of mutant cells out of the total number of cells corresponding to diploid and tetraploid across clones in TP53^−/− 184-hTERT sample, SA906a. This variant is present in both diploid and tetraploid cells of clone A. j, Same as i but for m.6869C>T in TP53^−/− 184-hTERT sample, SA906a. This variant is present in both diploid and tetraploid cells of clone G. k, Same as i but for m.6708G>A in SPECTRUM-OV-081 sample. This variant is present in both diploid and tetraploid cells of clone C. l, Ranking of the weight variable that indicates the probability of the variant contributing to the clonal assignment is plotted in a descending order. The real variants were colored in red. m, Heatmap of heteroplasmy across clones determined from mtDNA variants. The mtDNA variants-based clonal labels are on the top.