Traditional epidemiological studies have established numerous observational associations between human behaviors and/or diseases. Yet the causality relationship for such associations, which is central to disease treatment and drug development, is largely unknown. Mendelian randomization (MR) is an analytical method that statistically infers causal relationships from an exposure to an outcome (disease). It uses genetic variants associated with the exposure as instrumental variables for that exposure and can effectively overcome bias caused by unmeasured confounding factors. With the fruitful findings from hundreds of genome-wide association studies being conducted to date, instrumental variables for a variety of exposure traits are available, making the MR analysis being increasingly used to visit causal relationships for plenty of associations being established by traditional epidemiological studies.
Despite fruitful causal relationships being established by the MR approach, the progress is limited. While MR offers an attractive solution to causal inference using observational data, violation of some assumptions of MR may invalidate the findings. Specifically, the issue of pleiotropy, among others, has received much attention of empirical applications and methodological development in the past years. Progress towards comparative performance of existing methods as well as novel methodological development is expected.
The application of MR has extended from epidemiological analyses to new scenarios, such as mRNA or protein level research, to study causal relationships between metabolic biomarkers or molecular phenotypes. Other more sophisticated scenarios include microbiota-oriented causal inference as well as drug target discovery. All these applications will deepen the understanding of the pathophysiological mechanisms behind health problems.
This Research Topic will focus on all MR related empirical and methodological studies. The scope may include but is not limited to:
• Novel statistical method development;
• Comparative studies of existing statistical methods;
• Empirical causal inference between traditional traits/diseases;
• Causal inference between novel data types, such as microbiome, mRNA, and protein data;
• Application of drug target discovery.
Traditional epidemiological studies have established numerous observational associations between human behaviors and/or diseases. Yet the causality relationship for such associations, which is central to disease treatment and drug development, is largely unknown. Mendelian randomization (MR) is an analytical method that statistically infers causal relationships from an exposure to an outcome (disease). It uses genetic variants associated with the exposure as instrumental variables for that exposure and can effectively overcome bias caused by unmeasured confounding factors. With the fruitful findings from hundreds of genome-wide association studies being conducted to date, instrumental variables for a variety of exposure traits are available, making the MR analysis being increasingly used to visit causal relationships for plenty of associations being established by traditional epidemiological studies.
Despite fruitful causal relationships being established by the MR approach, the progress is limited. While MR offers an attractive solution to causal inference using observational data, violation of some assumptions of MR may invalidate the findings. Specifically, the issue of pleiotropy, among others, has received much attention of empirical applications and methodological development in the past years. Progress towards comparative performance of existing methods as well as novel methodological development is expected.
The application of MR has extended from epidemiological analyses to new scenarios, such as mRNA or protein level research, to study causal relationships between metabolic biomarkers or molecular phenotypes. Other more sophisticated scenarios include microbiota-oriented causal inference as well as drug target discovery. All these applications will deepen the understanding of the pathophysiological mechanisms behind health problems.
This Research Topic will focus on all MR related empirical and methodological studies. The scope may include but is not limited to:
• Novel statistical method development;
• Comparative studies of existing statistical methods;
• Empirical causal inference between traditional traits/diseases;
• Causal inference between novel data types, such as microbiome, mRNA, and protein data;
• Application of drug target discovery.