paramlink (version 0.2-1)

likelihood: Pedigree likelihood

Description

Calculates the likelihood of a pedigree, given the recombination rate between the disease locus and one marker locus.

Usage

likelihood(x, marker=NULL, t=NULL, logbase=NULL, TR.MATR=NULL, 
           singleNum.geno=NULL)

Arguments

x

a linkdat object, with x$model different from NULL.

marker

an integer between 0 and x$nMark, indicating which marker to use in the calculation. If NULL, the marker is taken to be all zeroes.

t

a numeric of length 1 indicating the recombination rate between the disease and marker loci. To make biological sense this should be between 0 and 0.5.

logbase

a numeric, or NULL. If numeric, the loglikelihood is computed, with logbase as basis for the logarithm.

TR.MATR

an 10*10*10 array with haplotype transmission probabilities, for speedy computation with diallelic markers.

singleNum.geno

Mostly for internal use: a numeric vector containing genotypes coded as a single integer: 00 -> 0, 11 -> 1, 22 -> 2, 12/21 -> 3, 01/10 -> 4, 02/20 -> 5. If this is non-null the marker argument is ignored.

Value

• The likelihood of the pedigree with disease marker data, given the disease model and the recombination rate. If marker=NULL, the result is the likelihood of the pedigree at the disease locus only (regardless the value of t). If the parameter logbase is numeric, the output is log(likelihood, logbase).

Details

The likelihood is calculated using the Elston-Stewart algorithm.

See Also

lod

Examples

ped = data.frame(ID=1:4, FID=c(0,0,1,1), MID=c(0,0,2,2), SEX=c(1,2,1,1), 
                 AFF=c(2,1,2,2), M_A1=c(1,1,1,1), M_A2=c(2,1,2,2))

x = linkdat(ped, model = 1)

# the likelihood at the disease locus:
likelihood(x, marker=NULL, t=0)

lod1 = likelihood(x, marker=1, t=0, logbase=10) - 
       likelihood(x, marker=1, t=0.5, logbase=10)
lod2 = lod(x, markers=1, t=0)
# should be the same:
stopifnot(identical(lod1, as.numeric(lod2)))