AbstractRobust inference of causal relationships between gene expression and complex traits using Mendelian Randomization (MR) approaches is confounded by pleiotropy and linkage disequilibrium (LD) between gene expression quantitative loci (eQTLs). Here we propose a new MR method, MR-link, that accounts for unobserved pleiotropy and LD by leveraging information from individual-level data. In simulations, MR-link shows false positive rates close to expectation (median 0.05) and high power (up to 0.89), outperforming all other MR methods we tested, even when only one eQTL variant is present. Application of MR-link to low-density lipoprotein cholesterol (LDL-C) measurements in 12,449 individuals and eQTLs summary statistics from whole blood and liver identified 19 genes causally linked to LDL-C. These include the previously functionally validatedSORT1gene, and thePVRL2gene, located in theAPOElocus, for which a causal role in liver was yet unknown. Our results showcase the strength of MR-link for transcriptome-wide causal inferences.
Mendelian Randomization Analysis Support Causal Associations of HbA1c with Circulating Triglyceride, Total and Low-density Lipoprotein Cholesterol in a Chinese Population
