RNOmni (version 0.1.4)

FIINT: Fully Indirect-INT

Description

Two-stage regression procedure In the first stage, phenotype is regressed on covariates and adjustments for population structure to obtain residuals. In the second stage, INT-transformed residuals are regressed on genotype.

Usage

FIINT(y, G, X, S, M = T, c = 3/8)

Arguments

y

Numeric phenotype vector.

G

Snp by obs genotype matrix.

X

Obs by feature covariate matrix.

S

Obs by feature structure matrix.

M

Apply missingness filter? See missFilter.

c

Offset applied during rank-normalization. See rankNormal.

Value

A numeric vector of p-values assessing the null hypothesis of no genotypic effect. P-values are estimated using the Wald statistic, and correspond to the rows of G.

Details

Note that, in simulations, FIINT did not consistently provide valid inference. For a similar approach that did control the type I error, see PIINT.

Examples

# NOT RUN {
# FIINT against normal phenotype 
p = RNOmni::FIINT(y=RNOmni::Y[,1],G=RNOmni::G[1:10,],X=RNOmni::X,S=RNOmni::S);
# }