Conditional inference in cis-Mendelian randomization using weak genetic factors.

Area of Science:

• Genetics
• Biostatistics
• Pharmacology

Background:

• Mendelian randomization (MR) uses genetic variants as instrumental variables to infer causal relationships between exposures and outcomes.
• Cis-MR, utilizing variants near a drug's target gene, supports drug target validation.
• Challenges arise when using multiple correlated variants with weak effects on the exposure.

Purpose of the Study:

• To develop robust cis-MR methods for causal inference using correlated genetic variants with potentially weak effects.
• To enhance drug target validation by improving the reliability of MR analyses in specific genetic regions.

Main Methods:

• Factor analysis to reduce dimensionality of correlated genetic variants within a single gene region.
• Development of conditional testing approaches for cis-MR inference.
• Extension of identification-robust tests for estimated genetic factors as instruments.
• Proposal of a new test accounting for first-stage screening of genetic factors.

Main Results:

• The proposed methods provide robust causal effect estimations even with weak genetic associations.
• Empirical results offer genetic evidence supporting cholesterol-lowering drug targets for coronary heart disease prevention.
• Factor analysis effectively handles the structured nature of genetic correlations in cis-regions.

Conclusions:

• The developed cis-MR methods enhance the validity of drug target inference.
• These statistical advancements contribute to the genetic validation of therapeutic targets for cardiovascular diseases.
• The approach offers a more reliable way to assess drug efficacy through genetic epidemiology.