Abstract
Purpose :
Genetic testing can provide a more precise diagnosis in inherited retinal diseases (IRDs). However, nearly 50% of patients do not have a genotypic diagnosis after clinical exome sequencing. We demonstrate that bioinformatic targeted long-read sequencing allows genome level analysis of a panel of IRD disease-causing loci to provide rapid, phased data sets. This can identify non-coding and structural variants to provide definitive diagnosis in those with prior inconclusive clinical testing.
Methods :
High molecular weight DNA was isolated from the blood of 30 subjects for library preparation and DNA sequencing on the Oxford Nanopore Technologies MinION flow cell. A panel of 373 genes was prepared from the hg38 genome assembly coordinates with a 50-kilobase flanking sequence in each direction. A haplotype-aware genotyping pipeline provided ranked output variants from long-read data of each subject for analysis.
Results :
In total the panel covered about 54.3 Megabases equating to ~1.7% of the human genome and produced 25X coverage on average. This allowed variant confirmation from prior clinical sequencing. The added benefit of variant haplotyping from long-read data could explain the differing disease phenotypes in those with shared genetic etiology. Furthermore, we show that in 5 cases without previous genetic testing or those with incomplete diagnosis that long-read sequencing can provide complete molecular diagnosis. We also demonstrate that full genomic coverage with long-read data can identify deep intronic variants and large structural variants missed by standard testing. Finally, since native DNA is sequencing, we utilized the methylation signal to improve gene assemblies and properly haplotype difficult cases to provide a full molecular diagnosis.
Conclusions :
Targeted long-read panel sequencing focuses sequencing bandwidth for increased depth of coverage. More importantly, bioinformatic targeting eliminates the need for specialized library preparation to allow sequencing of native DNA. Using this approach we are able to validate clinical sequencing data and solve cases of missing heritability due to the ability to comprehensively sample the full scale of variants in IRD genes. Moreover we demonstrate that phased data sets can provide molecular diagnosis within hours of sequencing compared to a timeline of weeks for standard clinical testing.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.