CH in twin pairs. (A) Concordance for CH status in age-matched MZ (n = 33) and DZ (n = 24) twin pairs. Observed (obs) proportions are those identified in the cohort, and expected (exp) are those generated by random sample permutation. There was no significant difference when comparing (i) the observed distributions between MZ and DZ twins (P = .59) and (ii) the observed vs expected distributions within the MZ (P = 1) or DZ (P = .86) twin groups (Fisher exact test). (B) Change in VAF over time in the 3 twin pairs in which both individuals had CH and serial samples were available. Each box surrounded by a solid line represents a twin pair. The fourth box surrounded by a dashed line is data from the third twin pair with the y-scale magnified to the lower VAFs. In each box, change in VAF over time is represented by a solid line for 1 twin, and a dotted line for the other, with line color indicating which gene was mutated. (C-D) The variants identified in 2 MZ twin pairs in which both twins harbored identical nonsense mutations: KDM6A (NM_021140:c.C2074T:p.Q692X) (C) and DNMT3A (NM_175629:c.C1792T:p.R598X) (D). Each “triplet” represents the 3 nonreference bases at each genomic position, centered on the identified variant position (denoted as 0). For each alternate allele, the VAF of each of the 154 individuals in the cohort is plotted. The horizontal dashed line represents the lower limit of sensitivity of variant calling, with calls below this VAF considered error. MZ twins with identical mutations are plotted in red, all other individuals in black. ns, not significant.