compareEquivalences {bayesCGH}R Documentation

Bayesian comparison of proportions of chromosomal gain and loss between clinical sets.

Description

This function compares a set of equivalence relations which define hypotheses of similarity within and differences between equivalence classes. The function is intended for use on array CGH data in the form of a ClassifySegList object.

Arguments

seg Object of class ClassifySegList.
group.arrays A list object of which each item should be a numeric vector giving the indices of the arrays in 'seg' belonging to a particular clinical set.
group.names Names of the clinical sets defined by 'sel.groups'.
probe.classes A list object of which each item is a character vector that describes the different groups of classifications of the aCGH data. Unlikely to change from default.
classes A list object, of which each item is another list object that defines an equivalence relation on the set of sets defined by 'sel.groups' (see Details).
class.names Names for each equivalence relation defined in 'classes'
class.priors Priors for each equivalence relation defined in 'classes' (see Details).
X.distinct Should the X chromosome be omitted from estimation of the prior?
Y.distinct Should the Y chromosome be omitted from estimation of the prior?
same.priors Do all equivalence relations defined in 'classes' have the same prior Dirichlet distribution on the different classifications defined in 'probe.classes'?

Details

This function takes a number of subsets of the 'seg' ClassifySegList structure and simultaneously tests a number of hypotheses defined in terms of equivalence classes upon them.

The subsets of 'seg' to be used are defined by 'sel.groups'. This is a list object in which the nth item is a numerical vector giving the arrays belonging to the nth subset. Names of the subsets can be defined in 'group.namess', a string vector.

The 'classes' argument is a list object of which each item is another list object defining an equivalence relation on the sets defined in 'sel.groups'. Names for each equivalence relation can be defined in 'class.names', a string vector. Prior likelihoods of each equivalence relation are defined in 'class.priors'.

Value

List object containing

data Description of data submitted to compareEquivalences.
equivalence.classes Description of equivalence classes submitted to compareEquivalences.
probabilities Matrix of probabilities for each probe belonging to each equivalence class.

Author(s)

Thomas Hardcastle

References

Hardcastle 2008

See Also

ClassifySegList


[Package bayesCGH version 0.6 Index]