anal.leaps: Analyze a backcross by model selection methods

In kbroman/qtlsim: Software for Testing QTL Analysis via Simulations

Analyze a backcross by model selection methods

Description

Analyze a backcross using forward or backward selection or either of these followed by branch-and-bound.

Usage

anal.leaps(dat, method=c("forward","backward","forwback",
                         "forwardleap","backwardleap"),
           bic.mult=2.5, max.steps=30)

Arguments

dat	The data: a list with components geno (the genotype data, as a matrix) and pheno (the phenotype data, as a vector).
method	Model search approach to use: forward selection, backward elimination, forward selection followed by backward elimination, or forward or backward selection followed by branch-and-bound.
bic.mult	Multiplier for BIC, for choosing the size of the model.
max.steps	Maximum number of steps in forward selection.

Value

A list whose first component is a vector indicating the marker columns selected and second component is the corresponding BIC-delta value.

Author(s)

Karl W Broman, [e-mail]

References

Broman, K. W. (1997) Identifying quantitative trait loci in experimental crosses. PhD dissertation, Department of Statistics, University of California, Berkeley.

See Also

simbc, anal.all, anal.mcmc, perm

Examples

dat <- simbc(n.ind=100,qtl.eff=c(1,1,1,-1,1,1,1)*0.75)
results <- anal.leaps(dat,method="forwback",bic.mult=2.56)

kbroman/qtlsim documentation built on May 17, 2023, 11:53 p.m.