Contains tools for combining the results of multiple gene expression studies. metahdep accounts for both sampling and hierarchical dependence among studies, as well as fundamental covariate differences between studies. It first formats the data for meta-analysis, starting with the original .CEL files. A necessary step in formatting is the matching of probesets across array versions based on common gene content. metahdep requires the construction of a data.frame object mapping probeset IDs to common ‘new names’ for multiple array versions. Next, an effect size estimate (and associated variance estimate) representing the degree of differential expression for each probeset in each study is calculated. A common platform is not necessary to use the metahdep package, although users will need to define appropriate effect size estimates and should be aware of possible cross-platform inconsistencies.