RNOmni: Rank Normal Transformation Omnibus Test

Implementation of genetic association tests that incorporate the rank-based inverse normal transformation (INT). These tests are broadly indicated for traits with continuous residual distributions. In the presence of non-normal residuals, INT-based tests robustly control the type I error, whereas standard linear regression may not. Moreover, INT-based tests dominate standard linear regression in terms of power. There are two main strategies for incorporating the INT in association analysis. In direct INT (D-INT), the trait is directly transformed. In indirect INT (I-INT), residuals are formed prior to transformation. Neither D-INT nor I-INT is uniformly most powerful. The INT omnibus test (O-INT) adaptively combines D-INT and I-INT into a single robust and statistically powerful approach.

Version: 0.6.0
Depends: R (≥ 3.2.2)
Imports: abind, foreach, plyr, Rcpp
LinkingTo: Rcpp, RcppEigen
Suggests: cowplot, ggplot2, knitr, reshape2, rmarkdown
Published: 2019-01-12
Author: Zachary McCaw [aut, cre]
Maintainer: Zachary McCaw <zmccaw at g.harvard.edu>
License: GPL-3
NeedsCompilation: yes
CRAN checks: RNOmni results

Downloads:

Reference manual: RNOmni.pdf
Vignettes: Rank Normal Omnibus Association Test
Package source: RNOmni_0.6.0.tar.gz
Windows binaries: r-devel: RNOmni_0.6.0.zip, r-release: RNOmni_0.6.0.zip, r-oldrel: RNOmni_0.6.0.zip
OS X binaries: r-release: RNOmni_0.6.0.tgz, r-oldrel: RNOmni_0.5.0.tgz
Old sources: RNOmni archive

Linking:

Please use the canonical form https://CRAN.R-project.org/package=RNOmni to link to this page.