Mapping structural variants to regulatory elements and disease genes associated with obesity

Abstract

Obesity is a complex condition with a significant genetic component. Genetic variants associated with obesity have been identified throughout the genome; however, the impact of these variants, especially in noncoding and regulatory regions, remains unclear. Recent advancements in long-read sequencing technologies have significantly improved the characterization of human genomes. In this study we used long-read sequences to better characterize complex areas of the genome such as noncoding regions that intersect regulatory elements. Fine-mapping structural variants to these regions was performed to identify candidate variants implicated in polygenic obesity.

We evaluated long-read whole genome sequences from 497 probands from the Genomic Answers for Kids cohort at Children’s Mercy Hospital. These probands were grouped into obesity cases and controls to characterize variants found in individuals with obesity compared with non-disease background. Rare structural variants were identified and mapped to trait-relevant genes from Polygenic Score Catalog. Enrichment of genes intersecting rare structural variants identified a gene set implicated in biological pathways for metabolism and neural signaling. Genomic constraint scores were applied to rare structural variant intersections revealing significantly higher constraint of structural variants in cases than controls. Duplications and deletions were found to have significantly higher constraint in cases in upstream regions, and gene bodies. CTCF-bound promoters were highly constrained in cases when intersected with deletions, but more constrained in controls when intersected with duplications. Constraint score analysis and gene enrichment identified many candidate variants associated with cases. These candidate variants have a high likelihood of being pathogenic making them novel targets for diagnostic and therapeutic efforts.