Allele Specific Information in Mendelian Randomization

• Activity Number: 535 - Contributed Poster Presentations: Section on Statistics in Genomics and Genetics
• Type: Contributed
• Date/Time: Wednesday, August 1, 2018 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Xuran Wang* and Nancy Zhang and Dylan Small and Mingyao Li
• Companies: University of Pennsylvania and University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
• Keywords: Mendelian randomization; Instrumental variable; allele specific expression; causal effect; expression quantitative trial loci

Abstract

Mendelian randomization (MR) is a statistical framework for quantifying the causal effect of genetic-modifiable exposure to phenotypes, in which genetic variants are used as instrumental variables (IV). With high throughput sequencing technologies, the exposure of interest can now encompass gene expression through the use of expression quantitative trail loci (eQTLs) as the IV and quantification of allele-specific gene expression became available. We present a new framework that makes use of allele specific expression to estimate the causal effect of gene expression on quantitative phenotypes. We call this new framework allele specific Mendelian randomization (ASMR). Simulations show that ASMR gives unbiased estimate of causal effect size, and, in many cases, improves estimation precision over two stage least square (TSLS). By utilizing allele-specific information, ASMR reduces estimation variance and improves detection sensitivity in cases where the eQTL strength varies across subjects. We illustrate the new framework on the problem of quantifying downstream effects of lncRNA expression on mRNA expression of protein coding genes on data from the GEUVADIS consortium.