likelihood: Pedigree likelihood
In pedprobr: Probability Computations on Pedigrees

Description Usage Arguments Details Value Author(s) References Examples

View source: R/likelihood.R

The likelihood() and likelihood2() functions constitute the heart of pedprobr. The former computes the pedigree likelihood for each indicated marker. The latter computer the likelihood for a pair of linked markers separated by a given recombination rate.

likelihood(x, ...)

## S3 method for class 'ped'
likelihood(
  x,
  markers = NULL,
  peelOrder = NULL,
  eliminate = 0,
  logbase = NULL,
  loop_breakers = NULL,
  verbose = FALSE,
  theta = 0,
  ...
)

## S3 method for class 'list'
likelihood(x, markers = NULL, logbase = NULL, ...)

likelihood2(x, ...)

## S3 method for class 'ped'
likelihood2(
  x,
  marker1,
  marker2,
  rho,
  peelOrder = NULL,
  eliminate = 0,
  logbase = NULL,
  loop_breakers = NULL,
  verbose = FALSE,
  ...
)

## S3 method for class 'list'
likelihood2(x, marker1, marker2, logbase = NULL, ...)

`x`	A `ped` object, a `singleton` object, or a list of such objects.
`...`	Further arguments.
`markers`	One or several markers compatible with `x`. Several input forms are possible: A `marker()` object compatible with `x`. A list of marker objects. A vector of names or indices of markers attached to `x`. If `x` is a list, this is the only valid input.
`peelOrder`	For internal use.
`eliminate`	Mostly for internal use: a non-negative integer indicating the number of iterations in the internal genotype-compatibility algorithm. Positive values can save time if the number of alleles is large.
`logbase`	Either NULL (default) or a positive number indicating the basis for logarithmic output. Typical values are `exp(1)` and 10.
`loop_breakers`	A vector of ID labels indicating loop breakers. If NULL (default), automatic selection of loop breakers will be performed. See `breakLoops()`.
`verbose`	A logical.
`theta`	Theta correction.
`marker1, marker2`	Single markers compatible with `x`.
`rho`	The recombination rate between `marker1` and `marker2`. To make biological sense `rho` should be between 0 and 0.5.

The implementation is based on the peeling algorithm of Elston and Stewart (1971). A variety of situations are covered; see the Examples section for some demonstrations.

complex inbred pedigrees
pedigrees with inbred founders
autosomal and X-linked markers
a single marker or two linked markers
markers with mutation models

A numeric with the same length as the number of markers indicated by markers. If logbase is a positive number, the output is log(likelihood, logbase).

Magnus Dehli Vigeland

Elston and Stewart (1971). A General Model for the Genetic Analysis of Pedigree Data. doi: https://doi.org/10.1159/000152448

### Example 1: Likelihood of trio with inbred father

x = cousinPed(0, child = TRUE)
x = addSon(x, 5)
x = relabel(x, old = 5:7, new = c("father", "mother", "child"))

# Equifrequent SNP marker: father homozygous, child heterozygous
m = marker(x, father = 1, child = 1:2)
x = addMarkers(x, m)

# Plot with genotypes
plot(x, marker = 1)

# Compute the likelihood
lik1 = likelihood(x, markers = 1)


### Example 2: Same as above, but using founder inbreeding

# Extract the trio
y = subset(x, c("father", "mother", "child"))

# Indicate that the father has inbreeding coefficient 1/4
founderInbreeding(y, "father") = 1/4

# Plot (notice the inbreeding coefficient)
plot(y, marker = 1)

# Likelihood should be the same as above
lik2 = likelihood(y, markers = 1)

stopifnot(all.equal(lik1, lik2))