Suzanne de Bruijn

247 Exploring the missing heritability in subjects with hearing loss and EVA region. After careful interrogation of the SVs, all of them were considered false positives based on SV length, and presence of the SVs in an in-house control dataset. The CEVA haplotype region was also manually inspected in the IGV software, which did not reveal any indications for potential SVs ( Figure S3B ). Interestingly, the insertion event that was detected with optical genome mapping and located just outside the CEVA region was also present in the long-read sequencing data (chr7:107,370,573, 1,612 bp insertion). This insertion was also present in available in-house control sequencing data, supporting the hypothesis the variant concerns a reference problem and is not a true SV. A comparable severity of hearing loss in the M1/CEVA and M2 cohorts As the CEVA haplotype was reported to be associated with a less severe HL phenotype than variants in the protein-coding or splice site regions of SLC26A4 23 , we addressed genotype-phenotype correlations in our cohort. We were able to retrieve pure tone audiometry of all subjects except for SLC071; for this subject, complete audiometry from only one ear was available ( Figure S2). The original CT or MRI scans of subjects SLC018 and SLC032 could not be retrieved. However, written reports of the imaging were available. Data on thyroid gland function were not consistently available and were therefore not included in this study. We applied the methods of Chao et al. to compare the severity of HL between four subject groups (M0, M1, M1/CEVA, and M2 Figure 3, Table 4 ). 23 The M1/CEVA group includes the M1/V1-CEVA subjects. Bilateral EVA was present in 7 of 10 (70%) M1/CEVA subjects, in all 4 M1 subjects without the CEVA haplotype, and 7 of 10 (70%) M0 subjects without the CEVA haplotype. All 11 M2 subjects (reference cohort, SLC048 and SLC085) had bilateral EVA. The median pure tone audiometry in the M2 group (85 dB HL, n = 20) was not significantly different from that of the M1/CEVA group (84 dB HL, n = 16) and the M1 group (79.5 dB HL, n = 8) (p-values 0.8300 and 0.7142, respectively, all adjusted for age). Also, no difference was observed between the M1/CEVA group and the M1 group (p = 0.8782). In contrast, when we compare the M2 and M1/CEVA groups with the M0 group, we observed significant differences in the severity of HL (p = 0.0015 and p = 0.0135, respectively). When compared to Chao et al, subjects in our study displayed a similar degree of median HL in the M2 group (86.3 and 85 dB in (23) and the present study, respectively), more severe HL in the M1/CEVA group (47.5 and 84 dB, respectively) and less severe HL in the M0 group (54.4 and 42 dB HL, respectively). Slight age differences were seen between the groups presented in Chao et al. and those in the current study ( Table S8 ). Chao and co- workers did not report audiometric data for the M1 group without the CEVA haplotype in trans , presumably due to the small sample size. Overall, in contrast to the study by