RMR-ICP: robust Mendelian randomization method accounting for idiosyncratic and correlated pleiotropy with applications to stroke outcomes.

Abstract

Motivation: Mendelian randomization (MR) serves as a valuable tool for investigating causal relationships between exposures and disease outcomes in observational studies. However, MR methods, operating under classical assumptions, may yield biased estimates and inflated false-positive causal relationships when faced with realistic and complex correlated horizontal pleiotropy (CHP). While numerous MR methods have emerged to address CHP effects, limited methods can effectively handle relatively large direct effects, commonly known as idiosyncratic pleiotropy.

Results: To address this gap, we propose an efficient and robust MR method to account for idiosyncratic and correlated pleiotropy, named RMR-ICP. Furthermore, Our method incorporates linkage disequilibrium structure using paralleled Gibbs sampling to enhance statistical power. The robustness and efficiency of our method are demonstrated through extensive simulation studies and applications. RMR-ICP is first used to analyze the effects of plasma proteins on stroke, followed by its application to conventional stroke risk factors. Our analysis reveals that Selectin E (SELE) exhibits a positive causal effect on the occurrence of any stroke. Only those specifically designed to account for idiosyncratic and CHP identified a significant positive causal effect of myeloperoxidase on ischemic stroke, with RMR-ICP providing stronger statistical evidence. Elevated Natriuretic Peptide B (BNP) levels specifically increase the risk of cardioembolic stroke (CES), though not with other stroke subtypes. This finding is consistent with previous studies suggesting that plasma BNP levels may help distinguish CES from other stroke types. Higher Waist-hip ratio (WHR) levels raise the risk across all stroke types. These findings provide new insights into identifying stroke-related risk factors.

Availability and implementation: RMR-ICP is publicly available at https://github.com/QingCheng0218/RMR.ICP.