Abstract
Purpose :
Short-read exome-based sequencing panels are most commonly used to genetically diagnose inherited retinal disease (IRD) patients. However, chromosomal maps for allelic variant discovery cannot be optimally reconstructed and up to 20-30% of autosomal recessive IRD patients have one or no identified disease-causing variants from short-read exome sequencing. Long-read genome sequencing strategies provide a more rapid and targeted approach for discovery of complex phased variants in IRD patients.
Methods :
Genomic DNA was isolated from blood of two siblings with clinical features consistent with Usher syndrome type 2 along with their unaffected mother. Whole genome short-read sequencing as well as targeted long-read sequencing with adaptive sampling methodology, which allowed us to focus our sequencing depth to just the USH2A locus, were carried out. A haplotype-aware genotyping pipeline and heuristic ranking of the deleteriousness of identified variants were performed from the sequencing reads. A variant context matching method was used to unravel any discordant variant calls between short- and long-read alignments.
Results :
Targeted long-read sequencing on the Oxford Nanopore Technologies platform provided phased variants of the three subjects. Our analysis demonstrated that ranked variants obtained from the haplotype-aware genotyping pipeline can focus on potential pathogenic candidates without a priori knowledge of the disease-causing variants. Moreover, sequencing data of the USH2A locus from targeted long-read data were able to identify all of the variants called by whole genome short-read data. The variants that were unique to long-read data were not called by short-read sequencing due to low coverage in those genomic regions, which often were located in non-coding portions of the gene. Application of a variant context matching tool to capture these perceived different calls showed that the two sequencing methods yielded highly concordant alignments, with a concordance of 99% for single nucleotide variants.
Conclusions :
A targeted long-read approach allows focused analysis of sequences of biologic interest and produces a rapid phased variant set with high recall and precision scores that can be used for clinical interpretation in IRD patients. Moreover, the phased data sets allow clinical narrowing of relevant disease-causing variants and allows diagnosis from the proband alone.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.