A robust and efficient method for Mendelian randomization with hundreds of genetic variants

Area of Science:

• Epidemiology
• Genetic Epidemiology
• Statistical Genetics

Background:

• Mendelian randomization (MR) is crucial for distinguishing correlation from causation in observational studies using genetic variants as instrumental variables (IVs).
• The validity of IVs is paramount for reliable MR findings, yet invalid IVs can bias results.
• Existing robust MR methods have limitations in handling invalid IVs and identifying distinct causal pathways.

Purpose of the Study:

• To develop a novel contamination mixture method for Mendelian randomization (MR).
• To enhance the robustness and efficiency of MR analyses in the presence of invalid instrumental variables.
• To identify distinct causal mechanisms by grouping genetic variants with similar causal estimates.

Main Methods:

• Developed a two-modality contamination mixture method for MR analysis.
• The method identifies clusters of genetic variants with similar causal estimates.
• Evaluated the method's performance against other robust MR techniques using mean squared error.

Main Results:

• The contamination mixture method demonstrated the lowest mean squared error across various realistic scenarios compared to existing robust methods.
• Identified 11 genetic variants associated with favorable lipid profiles (increased HDL-C, decreased triglycerides) and reduced coronary heart disease (CHD) risk.
• These variants showed consistent associations with blood cell traits, suggesting a shared mechanism linking lipids, blood cells, and CHD risk, potentially via platelet aggregation.

Conclusions:

• The contamination mixture method provides a robust and efficient approach for Mendelian randomization, particularly when dealing with invalid instrumental variables.
• The findings suggest a shared biological pathway linking lipid metabolism and coronary heart disease risk, mediated through platelet aggregation.
• This method advances causal inference in genetic epidemiology and offers insights into complex disease mechanisms.