Classical meta-analytic techniques may be applied to the problem of pooling evidence for linkage from different studies. The methodology relies on the availability of locus effect estimates (e.g. QTL for continuous traits, excess IBD sharing in Affected Relative Pairs designs) and associated standard errors on a common grid of chromosomal locations. We have implemented three models in R:
Homogeneity: same locus effect in all studies
- Size heterogeneity: the same locus is involved in all studies but with a possibly different effect size (due to e.g. allelic heterogeneity, varying allele frequency, interaction with other environmental factors)
- Locus heterogeneity: only a fraction of studies is linked to a locus
A test for heterogeneity is also implemented. In addition, we provide R functions that reads the output files obtained when running MERLIN . The tutorial should get you started quickly. All files needed (R functions, example files) along with a detailed description of the methodology can be downloaded here.
Citation: Heterogeneity in Meta-Analysis of QTL Studies by Hans van Houwelingen & Jérémie Lebrec. To appear in Meta-Analysis and Combining Information in Genetics. Edited by R. Guerra & D. Allison. CRC Press.