CorrMeta: Fast Association Analysis for eQTL and GWAS Data
with Related Samples and Correlated Phenotypes

Departments of Psychiatry, Biostatistics, Genetics and Computer Science
University of North Carolina at Chapel Hill

Center for Biomarker Research and Precision Medicine
Virginia Commonwealth University

Figure

Introduction

CorrMeta is a computationally efficient alternative to a linear mixed-effects model (LMM) for twin genome-wide association study (GWAS) or expression quantitative trait loci (eQTL) analysis. Instead of analyzing all twin samples together with LMM, CorrMeta first randomly splits twin samples into two independent groups on which multiple linear regression analysis is performed separately, followed by an appropriate meta-analysis to combine the two non-independent test results. Similar idea is also extended to combine GWAS results from multiple correlated phenotypes through CorrMeta. Our approaches provide a huge leap in terms of computing performance for GWAS data with related subjects and correlated phenotypes.

Key features

  1. Similar to meta-analysis, only summarized SNP level test statistics are necessary
  2. Fast alternative to linear mixed effect model with no inflation of type I error and negligible power loss
  3. Fast standard GWAS analysis for twin or correlated subjects
  4. Fast expression quantitative trait loci (eQTL) analysis for twin or correlated subjects
  5. Fast GWAS analysis for correlated phenotypes
  6. Implemented as an easy-to-use R package similar to MatrixEQTL

Download and document

R logo icon CorrMeta R package.
To install, download the package file CorrMeta_1.0.tar.gz and run (in R):

install.packages("MatrixEQTL")
install.packages("CorrMeta_1.0.tar.gz", repos = NULL, type="source")


The package includes reference manual and:
R logo icon Sample code
R logo icon Sample data set (file)


Reference

Xia, K, Shabalin, AA et al (2015). CorrMeta: Fast Association Analysis for eQTL and GWAS Data with Related Samples and Correlated Phenotypes. (Submitted)

Contact

Kai Xia: kxia@med.unc.edu

Andrey A Shabalin: ashabalin@vcu.edu

Fei Zou: fzou@bios.unc.edu