Extended Data Fig. 5: Evidence of multiple causal variants for 10q25.2 breakage and 1p CNN-LOH associations.

a, Multiple expanded repeats at FRA10B drive breakage at 10q25.2. We identified 12 distinct primary repeat motifs at FRA10B in 26 whole-genome-sequenced individuals from 14 families (labelled VNTR-N-x, where N denotes length in base pairs); carriers of these repeats exhibit varying degrees of FRA10B repeat expansion (Supplementary Note 8). The repeat motifs are AT-rich and are similar to FRA10B repeats previously reported³⁵. The alignment provided here includes the repeat motifs that were most frequently observed in FRA10B expanded alleles³⁵ (E8, E13, E17, and E19) along with a few other closely related expanded repeat motifs (E10, E11, and E12). b, Carriers of the 10q terminal deletion in the UK Biobank share long haplotypes at 10q25.2 identical-by-descent. Square nodes in the IBD graph correspond to males and circles to females. Node size is proportional to cell fraction and edge weight increases with IBD length. Coloured nodes indicate imputed carriers of variable number tandem repeats (VNTRs) at FRA10B (Supplementary Table 7); colour intensity scales with imputed dosage. c, Identity-by-descent graph at MPL locus (chr1:43.8 Mb) on individuals with mCAs on chr1 extending to the p telomere. Colored nodes indicate imputed carriers of SNPs independently associated with mosaic 1p CNN-LOH (Fig. 4a).