Enter Note Done


Keywords: Genome sequencing; Genome-wide association; Genomics; Mendelian disorder; Identification of disease genes

J. Krier 1; D. Hui 2; N. Patsopoulos 2

1) Medicine, Brigham & Women's Hospital, Boston, Massachusetts.; 2) Neurology, Brigham & Women's Hospital, Boston, Massachusetts.

Introduction. Many research initiatives, such as the Undiagnosed Disease Network (UDN), undertake exploratory analyses of genomic sequencing data to identify novel monogenic etiologies for rare disease. Key questions typically unexplored by these analyses, however, include whether the patient’s phenotype can be explained by an extreme polygenic risk burden of a related common disease and whether extreme polygenic risk burden to a common disorder can impact the expressivity of an underlying monogenic disorder. Here, we utilize genetic findings of well-characterized common diseases to identify patients with extreme polygenic risk burdens in a cohort of patients with suspected monogenic disease. Methods. Patients enrolled in Brigham Genomic Medicine/UDN Harvard Clinical Site received whole-genome sequencing for the affected proband and unaffected relatives. We generated genetic risk scores (GRSs) across multiple phenotypes and specifically focused on each proband’s most analogous common phenotype (e.g. multiple sclerosis for CNS white matter disease) based on statistically significant variants from recent GWAS meta-analyses. We further leveraged genome-wide data to calculate polygenic risk scores (PRS). GRS and PRS were corrected for the underlying ancestry leveraging reference population data. Results. Generally, the affected probands and their unaffected relatives did not have high values either GRS or PRS for the respective common diseases (z-score < 2 for any comparison), suggesting that picking analogous diseases is error-prone in such complex patients and/or that there may be an undiscovered monogenic or oligogenic etiology for probands. However, samples with high GRS or PRS for common diseases often had phenotypic characteristics suggestive of the respective disease. For example, two probands and families with early onset multiple sclerosis (MS) had GRS and PRS near the underlying ancestral population mean, z-scores < |1|. When we surveyed the entire cohort for MS polygenic risk burden, the two patients with the highest burden for MS (z-score >2) had MRI-evident delayed myelination and T cell mediated central nervous disorder. Conclusion. We outline here a nascent methodology and report on initial results for evaluating extremes of polygenic risk in rare disease cohorts. Additional data will be presented from our broader rare disease cohort (>200 affected individuals) including summary statistics and illustrative examples.