diptavo / metamultiskat

Region-based meta analysis of multiple correlated phenotypes

About

MetaMultiSKAT provides a meta-analysis framework for performing rare-variant based tests of associations with multiple phenotypes across multiple studies. It is a kernel regression framework extension of MultiSKAT. This is an initial working version. A number of functionalities will be added soon.

MetaMultiSKAT v(1.0)

This package contains the R-codes/functions (including an example dataset) to carry out the MetaMultiSKAT tests. The package is still under developement. Any questions or bug-reports should be mailed to [email protected]

Installation

MetaMultiSKAT R-package can be installed from the downloaded gzipped tarball as R CMD INSTALL MetaMultiSKAT_1.0.tar.gz with the following packages pre-installed: SKAT, copula, nlme, MultiSKAT

Citation

The manuscript will be uploaded soon to biorXiv.

Workflow

An example workflow has been shown in the wiki page

Main Functions

Meta.MultiSKAT.base: Performs MetaMultiSKAT tests with a given Sigma_s, Sigma_P and Sigma_G structure

Meta.MultiSKAT.wResample: Performs MetaMultiSKAT tests with a given Sigma_s, Sigma_P and Sigma_G structure and generates null p-values to be used in omnibus tests

minP.Meta: Minimum p-value based aggregate tests using copula

Meta.Hom: Performs MetaMultiSKAT test of association against the alternative that all the studies have similar effects

Meta.Het: Performs MetaMultiSKAT test of association against the alternative that all the studies have heterogeneous/uncorrelated effects

Meta.Com: Combination test of Meta.Het and Meta.Hom