null: Fitting the null model

In FREGAT: Family REGional Association Tests

Description

Gives estimation of model parameters under the null hypothesis

Usage

null(formula, phenodata, kin = NULL, opt.method = 'optimize',
ih2 = 0.3, eps = 1.e-04)

Arguments

formula	referring to the column(s) in phenodata to be analyzed as outcome and, if needed, covariates.
phenodata	a data frame containing columns mentioned in formula: trait to analyze and, if needed, covariates. Individuals not measured for trait or covariates will be omitted.
kin	a square symmetric matrix giving the pairwise kinship coefficients between analyzed individuals. Under default kin = NULL all individuals will be considered as unrelated.
opt.method	optimization method, one of "optimize" (default), "optim" and "nlminb". Corresponding R functions will be used in optimization.
ih2	initial value for h2. Default = 0.3.
eps	epsilon (precision) value for optimization. Default = 1e-04.

Details

The function performs one-dimensional optimization for h2 with analytical calculations for the other parameters.

Value

A list with values:

h2	estimate of the heritability
total.var	estimate of the total variance
alpha	estimates of fixed effects of covariates
df	the gradient
logLH	the total log-likelihood
p.normality	p value of Shapiro-Wilk normality test for the null model residuals.

Examples

data(example.data)

## Run the null model:
nullmod <- null(trait ~ age + sex, phenodata, kin)

## SKAT with the null model object obtained in the first run:
out <- FFBSKAT(trait ~ age + sex, phenodata, genodata, kin, nullmod)

FREGAT documentation built on Jan. 15, 2018, 9:04 a.m.