Related Experiment Videos
A method for meta-analysis of molecular association studies
Ammarin Thakkinstian1, Patrick McElduff, Catherine D'Este
1Clinical Epidemiology Unit, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand. raatk@mahidol.ac.th
Statistics in Medicine
|November 30, 2004
Summary
This study introduces a novel meta-analysis method for population-based molecular association studies. It addresses Hardy-Weinberg equilibrium and gene effect pooling without assuming a genetic model.
Area of Science:
- Genetics
- Biostatistics
- Population Genetics
Background:
- Meta-analysis methodology for population-based molecular association studies is underdeveloped.
- Key challenges include testing Hardy-Weinberg equilibrium (HWE) and biologically informed pooling of gene effect results.
Purpose of the Study:
- To propose a comprehensive meta-analysis process for population-based molecular association studies.
- To address the neglect in methodology concerning HWE testing and gene effect pooling.
Main Methods:
- The proposed method involves checking HWE using chi-square goodness of fit, with sensitivity analysis.
- It includes heterogeneity checking and exploration of causes.
- Regression analysis is used for pooling data and determining gene effects if heterogeneity is absent.
Main Results:
- A step-by-step process is outlined for meta-analysis.
- The method allows data to determine the best genetic model, avoiding a priori assumptions.
- It is implementable using standard statistical software.
Conclusions:
- The developed meta-analysis approach provides a robust framework for population-based molecular association studies.
- It enhances the reliability of pooled results by addressing HWE and gene effect modeling.
- This methodology offers flexibility by not requiring pre-defined genetic models.