Participants
Keith Zirkle
Meta-analysis of microarray studies using quality-adjusted random effect size model
With the rise of microarray studies in recent years, an enormous amount of data is available from multiple studies on the same phenomena. Most of these studies have low to moderate sample size. Systematic integration of such studies would allow the statistical power of detecting differentially expressed genes to increase. We study a meta-analysis technique based on traditional effect size approach with random effects model. In this study, effect-size for each gene is weighted by a quality measure reflecting the study quality. This quality weight increases consistency in identifying differential gene expression and decreases heterogeneous results between multiple datasets.