Structural variant calling with GATK-SV finds hidden complex structural variants in unresolved inherited retinal diseases

초록

Purpose : The diagnostic rate of exome or targeted panel sequencing for inherited retinal diseases (IRDs) is approximately 60%, and genome sequencing (GS) does not bring substantial improvements. This study incorporates an integrative analysis using GATK-SV to uncover hidden genetic causes in a multinational cohort.Methods : We analyzed 214 unresolved IRD probands and their available family members (n=266) using GS from two centers (Massachusetts Eye and Ear, USA and Severance Hospital, Korea). A comprehensive bioinformatic pipelines, including GATK-SV (Manta, Wham, GATK gCNV, cn.Mops and SCRAMble) for structural variant (SV) detection, ExpansionHunter for trinucleotide repeanalysis, and MitoSAlt for mitochondrial SV detection, were applied. Results : Of the 214 probands, 114 (53.3%) were female, and most of them were non-Finnish European (48.1%), East Asian (39.7%), or African/African-American (5.1%). The diagnostic yield after GS was 97/214 (45.3%). Of the 97 families with identified causative variants, 32 (33.0%) required GS for identification of variants including deep intronic variants (n=14), copy number variations less than 3 exons (n=12), regulatory SVs (n=3), mitochondrial SV (n=1), copy-neutral inversions (n=1), or complex rearrangement (n=1). The remaining 65 families (67.0%) had variants that should have been identified by the prior exome sequencing. The reason for prior missed diagnosis were mis-interpretation of coding variants such as common hypomorphic variants (n=42), regulatory variants that would have been covered by current exome capture (n=10), copy number variations detectable by exome (n=6), mitochondrial SNVs (n=3), variants in repetitive regions (n=2), or mobile element insertions in exons (n=2). Conclusions : Our study demonstrates that GS identified causative variants in nearly half of previously unresolved IRD cases. However, approximately two-thirds of these cases could be resolved with exome reanalysis or updated exome sequencing methods. The unique capabilities of GS including identification of deep intronic variants or copy-neutral SVs, accounted for about 15% of unresolved cases.