Genome-First Observational Study of Neuromuscular Genetic Variants Associated Phenotypic Spectrum in NIH All Of US Cohort

Topic:

Translational Research

Poster Number: LB436

Author(s):

Zhiyv Niu, PhD, FACMG, Mayo Clinic, Milovan Suvakov, PhD, Mayo Clinic, Emily C. Lauer, MS, Mayo Clinic, Angela M. Pickart, MS, Mayo Clinic, Jie Na, MS, Mayo Clinic, Duygu Selcen, MD, Mayo Clinic, Margherita Milone, MD, PhD, Mayo Clinic, Alexej Abyzov, PhD, Mayo Clinic, Christopher J. Klein, MD, Mayo Clinic

Background:
Traditional genetic testing uses a phenotype-first approach based on patient symptoms. However, many genetic diseases exhibit broad variability due to variant types, allelic expression, modifying factors, and age/sex influences. Advanced genome sequencing and EMR systems now enable a genome-first approach for investigating a broader phenotypic spectrum and atypical presentations. A deeper and more complete understanding of neuromuscular gene and genetic variant expressivity is critical in the era of RNA and gene therapy.
Objectives:
We conducted a genome-first study of neuromuscular diseases by integrating whole-genome (WGS) with EMR data on the NIH All of US Research Platform. The pilot focused on common neuromuscular genes, including PMP22 and DMD, to build a cloud analytics for small variant and copy number variant (CNV) leading to a phenome-wide association study (PheWAS).
Results:
A tiered CNV analysis was performed on 245,400 WGS samples from the AllofUS V7 cohort. CNV result with microarray confirmation was combined with short variant. CNVs containing full PMP22 gene were detected in over 150 participants, consistent with a ~1 in 2,000 frequency. PheWAS revealed distinct phenotypes between PMP22 duplication and deletion carriers, and an increased odds ratio for inflammatory neuropathy with the deletion. In contrast, numerous CNVs in DMD coding region were identified, including truncating and many variants of uncertain significance (VUS). PheWAS in adult males with DMD truncating variants suggested that variants affecting different DMD protein regions can result in milder phenotypes versus severe Duchenne muscular dystrophy. This approach provides a framework for evaluating CNV VUS using EMR and survey data.
Conclusion:
This genome-first observational study, on key neuromuscular disease genes confirms significant variations in PMP22 and DMD related neuromuscular phenotypes. It demonstrated the feasibility of EMR aided genetic expressivity and variant classification study. aid . We also identified challenges of future genome-first precision medicine on informatics, data strategies, and decision support.