Improving Rare Disease Diagnosis With Advanced Genetics and Long Read Sequencing

Mendelian disease affects approximately 1 in 17 people across the globe and has so far been associated with
over 4800 genes. There is a compelling need to improve our understanding of rare genomic variation; current
genetic analysis techniques for Mendelian disease have a limited success rate meaning that despite
comprehensive genomic testing, many patients and families are left without a molecular diagnosis. It is a pivotal
time for rare disease research, diagnostics and genomic medicine, with whole genome sequencing (WGS) likely
to be the method of choice for genetic analysis for the short to mid-term future. However, there are significant
limitations with the current chemistry and informatics technology, and our understanding/interpretation of
variants; Next Generation Sequencing (NGS) relies on short, paired-end read sequencing of up to 150bp which
comprise limited genomic context, and to truly capitalize on WGS coverage, we must better understand the
variation outside of gene panels and coding exons that remain the focus of clinical diagnostics. Thus, there are
many contributors to the missing heritability including undiscovered genes, variants of uncertain significance
(VUS), non-exonic variants and structural rearrangements and genes/variants intractable to NGS pipelines. The
primary goal of this study is to better understand missing heritability in rare disease. Previous work in my
laboratory has demonstrated that novel genes and non-coding variants in known inherited retinal dystrophy (IRD)
genes are a significant cause of disease and we have established methodology to be able to characterize the
effect of splice variants using blood derived mRNA and nanopore sequencing. Simultaneous work on the utility
of ultralong-read sequencing for difficult to resolve cases has shown promising preliminary results. This
application aims to build on these findings and develop the studies in IRD genes, expanding to broader genetic
disease patients and families who have undergone testing at GGC. Aim 1 will utilize existing anonymized WGS
datasets from the 100,000 genomes project to identify novel candidate disease genes and pathogenic noncoding variants by applying cutting-edge bioinformatics tools. Up to two genes will be taken forward in functional
studies in collaboration with other research groups as separate projects. Noncoding/VUS will be analyzed by
RT-PCR/nanopore sequencing to determine damaging effects and thus reclassify those variants as pathogenic.
Aim 2 will investigate the potential utility of adaptive sampling targeted nanopore sequencing for clinical use in
unsolved patients and families, and those where it is suspected that the culprit gene is intractable to NGS. For
this exploratory work, we will use the model of IRD to test this in the first instance, targeting genes including
OPN1LW/OPN1MW, ABCA4, USH2A, EYS, PRPF31, TYR, genes known to have limitations with coverage,
phasing or haplotypes that limit the ability of NGS for diagnosis. Further targets will be investigated in
collaboration with GGC leadership. We will examine the ability of adaptive sampling long-read sequencing to
provide sufficient coverage, depth and read length to enable full characterization of clinically relevant intractable
genomic regions. These studies are novel, being at the cutting edge of genomics for human disease and
capitalize on the expertise of the applicant and the outstanding research environment of GGC.

Grant Number: 5P20GM139769-05
NIH Institute/Center: NIH
Principal Investigator: Gavin Arno