Abstract:
In this talk I’ll presents methods for detecting, quantifying, and interpreting dominance, recessivity, and compound heterozygosity across population-scale cohorts. First, I introduce dominance LD score regression (d-LDSC), to estimate genome-wide “dominance heritability” from summary statistics while accommodating site-specific nonadditive effects. To boost power to detect recessive effects driven by rare variants, I present a framework that incorporates phase information to infer compound-heterozygous variants in exome data. In 175,587 UKB participants, this increases identification of damaging biallelic genotypes, and reveals recessive gene–phenotype associations, including bi-allelic ATP2C2 associated with earlier COPD onset. I further describe a federated recessive meta-analysis across six biobanks. Finally, I introduce an orthogonal recoding framework that plugs in to existing software to deviations from an additive model at the gene-level.

Short bio:

I am a SMARTbiomed Senior Research Fellow, based at the BDI and the Department of Statistics. I co-lead the Biobank Rare-Variant Consortium (BRaVa), a collaborative effort to analyse rare genetic variation at scale to identify the genetic underpinnings of complex traits and disorders. I also co-lead the (gen)omics theme at the BDI. I am particularly interested in combining genetic, phenotypic and functional data to better refine association signals and understand the genetic architecture of complex traits.

Join Zoom Meeting

aarhusuniversity.zoom.us/j/63936720258