mpprob: Compute founder probabilities for multi-parent crosses
In behuang/mpMap: Multi-Parent RIL Genetic Analysis

Description Usage Arguments Details Value See Also Examples

Using haplotype probabilities, computes the probability that each location on a genome was inherited from each founder. Locations are run either at markers only, at the midpoints of all intervals or at step sizes of x cM. Probabilities can be computed using internally, or with R/happy.hbrem or R/qtl.

mpprob(object, chr, step = 0, mrkpos = TRUE, mapfx = c("haldane",
  "kosambi"), ibd = FALSE, threshold = 0.7, program = c("mpMap", "qtl",
  "happy"), tempfiledirectory = "", generations = 5, est = TRUE,
  geprob = 1e-04)

`object`	Object of class `mpcross`
`chr`	Subset of chromosomes
`step`	Step size (in cM) to create grid of positions at which to compute probabilities. At default value of 0, probabilities are calculated at marker positions only
`mrkpos`	Flag to indicate whether to compute probabilities at both marker positions and step size or just step size. Is overridden for step size of 0.
`mapfx`	Map function used to convert map to recombination fractions
`ibd`	Flag to indicate whether to compute probabilities using IBD genotypes
`threshold`	Threshold for calling founder probabilities
`program`	R package to use to compute probabilities
`tempfiledirectory`	Directory in which to output temporary files. Default is current working directory
`generations`	Number of generations to assume in HAPPY. see `happy`
`est`	Flag for whether to impute founder alleles
`geprob`	Probability of genotyping error

If program=="mpMap" then probabilities are computed using flanking markers at positions across the genome and represent 3-point haplotype probabilities. If program=="happy" then probabilities are computed using default values in R/happy.hbrem, which calculates ancestral haplotypes without using pedigree information. This only allows for probabilities to be computed at midpoints of intervals. If program=="qtl" then probabilities are computed from multipoint founder probabilities in R/qtl. The default is to use qtl.

If step==0 for R/happy.hbrem, then probabilities are computed at the midpoints of marker intervals. However, if step==0 for R/qtl or R/mpMap, probabilities are computed only at marker locations.

The input mpcross object is returned with two additional components:

`prob`	A list with founder probabilities for each chromosome. Format is a matrix with n.founders * n.markers columns and n.lines rows. Each group of n.founders columns will add up to 1. Founder probabilities are in the order of founders in the input founder matrix.
`estfnd`	A list with estimated founders for each chromosome. Format is a matrix with n.markers columns and n.lines rows. Missing values indicate where no founder probability exceeded the input threshold. Numeric values for founders indicate the row in the input founder matrix corresponding to the estimated founder.

plot.mpprob, summary.mpprob

sim.map <- qtl::sim.map(len=rep(100, 2), n.mar=11, include.x=FALSE, eq.spacing=TRUE)
sim.ped <- sim.mpped(4, 1, 500, 6, 1)
sim.dat <- sim.mpcross(map=sim.map, pedigree=sim.ped, 
	qtl=matrix(data=c(1, 50, .4, 0, 0, 0), 
	nrow=1, ncol=6, byrow=TRUE), seed=1)
mpp.dat <- mpprob(sim.dat, program="qtl")
plot(mpp.dat)
summary(mpp.dat)