multic (version 0.3.2)

addGE: Assess combinations of univariate multic objects

Description

Determine whether there is any evidence that running a multivariate multic model will significantly increase the evidence of a genetic effect.

Usage

addGE(multic.objs, combine=2, plotit=FALSE, ibd.dist, statistic=c("lrt",
"wald"), legend=TRUE, ...)

Arguments

multic.objs

A list of 1-trait multic objects.

combine

Indicate how many traits should be examined together. The program will then look at all N traits choose 'combine'.

plotit

Logical, default=FALSE. If TRUE, a LOD plot is generated with a separate line for each combination of traits.

ibd.dist

The default is to use the distances from the first multic object. This options allows the user to provide a different set of distances.

statistic

Character, default="lrt". This determines whether the Wald statistic (MG/SE) or the LRT is used when combining the traits.

legend

Logical, default=TRUE. If TRUE and if plotit=TRUE then a legend is automatically provided.

...

Allows for graphical parameters to be passed to the plot function (only applicable when plotit=TRUE).

Value

• A data frame is returned if the function is assigned to new object. Included are the various combinations (listed in order 1-N), the Chi-square statistic, the p-value, the distance, and the LOD score.

Side Effects

If plotit=T, a plot is generated on the current graphical device.

References

M. de Andrade, C. Olswold, J.P. Slusser, L.A.Tordsen, E.J. Atkinson, K.G. Rabe, and S.L.Slager. Identification of genes involved in alcohol consumption and cigarette smoking. BMC Genetics, 6:S112, 2005.

See Also

multic, gene.eff

Examples

add2 <- addGE(list(bmi10, dia10, sys10), combine = 2, plotit = T, ylim=c(0,8), legend=F)
add3 <- addGE(list(bmi10, dia10, sys10), combine = 3, plotit = F)
lines(add3$cM, add3$lod, col=4, lwd=2, lty=4)
key(corner=c(0,1), lines=list(lwd=2, col=1:4, lty=1:4),
    text=list(c('BMI-Dia','BMI-Sys','Dia-Sys','BMI-Dia-Sys'), col=1:4))