Suzanne de Bruijn

246 Chapter 5 pathogenic variant in SLC26A4 (M1/CEVA). All heterozygous variants with an AF ≤5% in gnomAD that were shared between the two individuals and located within the determined boundaries of the CEVA haplotype were analyzed ( Table S7 ). In total, 20 shared variants remained and included the 12 original SNPs that previously defined the CEVA haplotype. 22 Sixteen of the shared variants are located in intronic regions, but for none of them, a significant effect (score ≥0.1) on splicing is predicted by SpliceAI. Two variants are located within a cis regulatory element of SLC26A4 according to the GeneHancer database, however, these also show overlap with a long interspersed nuclear element (LINE) repeat element. One variant (CEVA SNP9) has a high nucleotide evolutionary conservation score (PhyloP, 2.769 [range -14, 3]). No SVs or CNVs were detected within or overlapping with the CEVA haplotype and shared by the two individuals. Regions harboring heterozygous variants with an AF ≤5% in gnomAD that were not shared between SLC012 and SLC036 had sufficient coverage to exclude that these variants were only called in one of the subjects but present in both of them. None of the variants identified in either SLC012 or SLC036 were within the SLC26A4 gene or were obviously deleterious. SVs and CNVs within the CEVA boundaries were analyzed separately for the two subjects which did not reveal any of such variants that were not shared by the two studied subjects. To fully exclude that the CEVA haplotype harbors different pathogenic variants in the studied individuals, a study design including short- and long-readWGS in several nuclear families has to be applied. Optical genome mapping & long-read sequencing To investigate the possibility that SVs were missed using short-read sequencing, optical genome mapping (Bionano Genomics) was performed using ultra-high molecular weight DNA isolated from peripheral blood cells of individual SLC012 (M1/CEVA). Optical genome mapping identified a total of 6,565 SVs, of which none were within the CEVA region (D7S501-D7S2459; chr7:106,440,266-107,360,254). Two SV calls (both calling the same 2,196 bp insertion between chr7:107,367,549 and 107,373,585) were located just outside this region ( Figure S3A ). This insertion was also called in 100% of our current optical genome mapping control cohort 71 , strongly suggesting that this reflects a reference problem rather than a real SV. Additionally, there were 22 CNV calls, of which none were within the CEVA region. Subsequently, PacBio long-read sequencing was performed on genomic DNA isolated from individual SLC079 (M1/CEVA; in trans status unknown). SV analysis of the long- read sequencing data revealed a total of 55,205 SVs, of which 12 within the CEVA