mppR (version 1.5.0)

parent_cluster.mppData: Parent clustering for mppData objects

Description

Integrate the parent clustering information to the mppData object. The parent clustering is necessary to compute the ancestral model. If the parent clustering step is skipped, the ancestral model can not be used but the other models (cross-specific, parental, and bi-allelic) can still be computed.

Usage

parent_cluster.mppData(mppData, par.clu = NULL)

Value

An increased mppData object containing the the same elements as the mppData object provided as argument and the following new elements:

par.clu

Integer matrix with rows representing markers and columns corresponding to the parents. At a single marker position, parents with the same value were clustered in the same ancestral group.

n.anc

Average number of ancestral clusters along the genome.

mono.anc

Positions for which the ancestral clustering was monomorphic.

Arguments

mppData

An object of class mppData. the mppData must have been processed using: create.mppData, QC.mppData, IBS.mppData, and IBD.mppData.

par.clu

Interger matrix representing the results of a parents genotypes clustering. The columns represent the parental lines and the rows the markers. The columns names must be the same as the parents list of the mppData object. The rownames must be the same as the map marker list of the mppData object. At a particular position, parents with the same value are assumed to inherit from the same ancestor. for more details, see par_clu. Default = NULL.

Author

Vincent Garin

Details

At a single marker position, two parents can be grouped into a similar ancestral classes if we assume that they receive there allele from a common ancestor. The parent clustering information (par.clu) describe parental relatedness and which parent belong to which ancestral group. For example, at marker i, we could have five parents (pA, pB, pC, pD, pE) and the following clustering information (1, 2, 1, 2, 3). This means that pA and pC received their allele from the same ancestor (A1). pB and pD also have a shared ancestor (A2) who is different from (A1). And pE was not included in any group and can be seen as an independent ancestral group (A3).

The parent clustering information is provided via par.clu. It is an interger matrix with markers in row and parents in columns. At a particular marker position, parents with the same value are assumed to inherit from the same ancestor. for more details, see par_clu.

The marker positions that are considered as monomorphic given the parent clustering information are set back to one allele per parent to still allow the computation of the QTL allelic effect at those positions later.

The parent clustering can be performed using the R package 'clusthaplo' that can be found there: https://cran.r-project.org/src/contrib/Archive/clusthaplo/. The 'clusthaplo' option is not integrated in this version of mppR. However, a version of mppR with function calling clusthaplo can be found on github https://github.com/vincentgarin/mppR (branch master).

See Also

create.mppData, QC.mppData, IBS.mppData, IBD.mppData, par_clu

Examples

data(mppData_init)
data(par_clu)

mppData <- QC.mppData(mppData_init)
mppData <- IBS.mppData(mppData = mppData)

mppData <- IBD.mppData(mppData = mppData, type = 'RIL',
                       type.mating = 'selfing')
                       
mppData <- parent_cluster.mppData(mppData = mppData, par.clu  = par_clu)