Evaluating the potential role of pleiotropy in Mendelian randomization studies

Area of Science:

• Genetics
• Epidemiology
• Bioinformatics

Background:

• Pleiotropy, a single genetic variant influencing multiple traits, is prevalent in the human genome.
• Mendelian randomization (MR) uses genetic variants to infer causal relationships between traits, assuming vertical pleiotropy.
• Horizontal pleiotropy, where variants affect traits via independent pathways, poses a significant challenge to MR validity.

Purpose of the Study:

• To review the challenges posed by horizontal pleiotropy in Mendelian randomization.
• To outline newly developed methods for addressing horizontal pleiotropy.
• To enhance the reliability of causal inference in phenome-wide association studies.

Main Methods:

• Review of existing and emerging statistical methods for Mendelian randomization.
• Focus on differentiating vertical from horizontal pleiotropy.
• Application of these methods to genome-wide association study data for causal inference.

Main Results:

• Genome-wide association studies provide vast genetic data for phenome-wide causal inference.
• Newly developed methods offer improved approaches to detect and correct for horizontal pleiotropy.
• These advancements increase the trustworthiness of causal estimates derived from MR.

Conclusions:

• Addressing horizontal pleiotropy is crucial for robust causal inference using Mendelian randomization.
• The integration of novel statistical techniques can significantly improve the reliability of MR studies.
• This review provides a framework for utilizing advanced methods to overcome pleiotropy challenges in genetic epidemiology.