mlreg.p: EXPERIMENTAL Linear and logistic regression and Cox models for genome-wide SNP data

Description

Linear and logistic regression and Cox models for genome-wide SNP data

Usage

mlreg.p(formula, data, snpsubset, idsubset, gtmode = "additive", trait.type = "guess")

Arguments

formula

Standard formula object

data

an object of gwaa.data-class

snpsubset

Index, character or logical vector with subset of SNPs to run analysis on. If missing, all SNPs from data are used for analysis.

idsubset

Index, character or logical vector with subset of IDs to run analysis on. If missing, all people from data/cc are used for analysis.

gtmode

Either "additive", "dominant", "recessive" or "overdominant". Specifies the analysis model.

trait.type

Either "gaussian", "binomial" or "survival", corresponding to analysis using linear regression, logistic regression, and Cox proportional hazards models, respectively. When default vale "guess" is used, the program tries to guess the type

Value

An object of scan.gwaa-class

Details

Linear regression is performed using standard approach; logisitc regression is implemented using IRLS; Cox model makes use of code contributed by Thomas Lumley (survival package).

For logistic and Cox, exp(effB) gives Odds Ratios and Hazard Ratios, respectively.

Examples

Run this code

	require(GenABEL.data)
	data(ge03d2)
	dta <- ge03d2[,1:100]
# analysis using linear model
	xq <- mlreg.p(bmi~sex,dta)
# logistic regression, type guessed automatically
	xb <- mlreg.p(dm2~sex,dta)
# Cox proportional hazards model, assuming that age is the follow-up time 
# generally this does not make sense (could be ok if age is age at onset)
	xs <- mlreg.p(GASurv(age,dm2)~sex,dta)