FunColoc: A Generalized Functional Regression Model for Genetic Colocalization Analysis of microRNA Counts and Disease-related Outcomes

Document Type

Conference Proceeding

Publication Date

10-1-2024

Publication Title

Genet Epidemiol

Abstract

MicroRNAs play an important role in regulating gene expression and are increasingly associated with complex diseases such as osteoarthritis. Colocalization analysis of microRNA counts and disease-related outcomes can reveal whether they share the same genetic association signal(s) in a locus, which can point to novel regulatory pathways. Existing colocalization methods do not take full advantage of individual-level data, where both traits are collected in the same set of individuals. In addition, these methods are not applicable to traits with different types of distributions (such as microRNA counts and a binary/quantitative disease-related outcome). To address these challenges, we propose FunColoc, a generalized multivariate Functional regression model for Colocalization analysis between microRNA counts and an outcome with similar or different types of distributions (e.g., binary/quantitative). FunColoc estimates continuous trait specific functions for the effects of multiple variants in a locus depending on their positions. We construct a colocalization test statistic defined as the product of the variant effect functions between the two traits. We then compute P-values for colocalization testing under a composite null hypothesis (i.e., both estimated variant effect functions are different from zero) using empirical approaches. Through simulation studies in a genomic locus randomly chosen on 16q in 5,000 individuals with simulated genotypes, microRNA counts and osteoarthritis severity, FunColoc detected colocalization with good power (60-80%) in various scenarios with colocalization and controlle Type-I errors (<5%) in scenarios without colocalization. Next steps include applying FunColoc to the Osteoarthritis Initiative cohort, a longitudinal observational knee osteoarthritis study.

Volume

48

Issue

7

First Page

349

Find It @ Sladen

Share

COinS