formetascore: function to run GWA analysis oriented for future meta-analysis

Description

Function to run GWA analysis -- using all functions available in GenABEL -- and produce output oriented for future meta-analysis

Usage

formetascore(formula, data, stat = qtscore, transform = "no", 
		build = "unknown", verbosity = 1, ...)

Arguments

formula

standard formula

data

object of gwaa.data-class

stat

which GWA analysis function to apply. Could be mlreg, qtscore, mmscore, grammar, egscore, etc.

transform

Which trait transform to apply, could be ztransform or rntransform. Default value is "no" -- no transformation.

build

if you need that in output, specify genomic build here (e.g. "35")

verbosity

how much output is produced? Possible values are 0, 1, and 2

...

further arguments, passed to the ''stat'' GWA analysis function

Value

Data frame, containing GWA summary. The fields include: (1) SNP name (2) chromosome (3) position (4) number of people with available data (5) effect of the allele (6) standard error of the effect (7) P-value for the test (8) corrected P-value (we will use Genomic Control) (9) coding, with reference allele coming first (10) strand (11) frequency of the reference allele (12) Exact P-value for HWE test, etc. (depends on "verbosity" parameter.

Details

This function should be applied to analysis of quantitative traits, if meta-analusis is aimed afterwards.

A transformation is applied to the formla-defined residual, and the resulting trait is analysed with specified function. Results are arranged as data-frame.

Examples

Run this code

	require(GenABEL.data)
	data(ge03d2c)
	x <- formetascore(bmi ~ sex+age,ge03d2c)
	x[1:10,]
	x <- formetascore(bmi ~ sex+age,ge03d2c,trans=ztransform)
	x[1:10,]
	x <- formetascore(bmi ~ sex+age,ge03d2c,trans=rntransform,verbosity=2)
	x[1:10,]