GetRelM: Matrix with Pairwise Relationships
In sequoia: Pedigree Inference from SNPs

GetRelM

R Documentation

Matrix with Pairwise Relationships

Description

Generate a matrix or 3D array with all pairwise relationships from a pedigree or dataframe with pairs.

Usage

GetRelM(
  Pedigree = NULL,
  Pairs = NULL,
  GenBack = 1,
  patmat = FALSE,
  Return = "Matrix",
  Pairs_suffix = "?"
)

Arguments

`Pedigree`	dataframe with columns id - dam - sire.
`Pairs`	dataframe with columns ID1 - ID2 - Rel, e.g. as returned by `GetMaybeRel`. Combining `Pedigree` and `Pairs` works best if the relationships are coded as listed below.
`GenBack`	number of generations back to consider; 1 returns parent-offspring and sibling relationships, 2 also returns grand-parental, avuncular and first cousins.
`patmat`	logical, distinguish between paternal versus maternal relative pairs? For avuncular pairs, the distinction is never made.
`Return`	'Matrix', 'Array', or 'List'. 'Matrix' returns an N x N matrix with the closest relationship between each pair. 'Array' returns an N x N x R array with for each of the R considered relationships whether it exists between the pair (1) or not (0). See Details below. 'List' returns a list with for each of the R considered relationships a 2-column matrix with the IDs of the pairs having such a relationship. The size of the list (in Mb) is much smaller than for the matrix or array, and this is therefore the only format suitable for pedigrees with many thousands of individuals. If `Pairs` is specified, the only possible return type is 'Matrix'.
`Pairs_suffix`	symbol added to the relationship abbreviations derived from `Pairs`, when both `Pedigree` and `Pairs` are provided. Can be an empty string.

Details

Double relationships are ignored when Return='Matrix', but not when Return='Array'. For example, when A and B are both mother-offspring and paternal siblings (A mated with her father to produce B), only the mother-offspring relationship will be indicated when Return='Matrix'.

Note that full siblings are the exception to this rule: in the Array they will be indicated as 'FS' only, and not as 'MHS' or 'PHS'. Similarly, full avuncular pairs are not indicated as 'HA'. Double half-avuncular relationships are indicated as both FA and HA.

When Pairs is provided, GenBack and patmat are ignored, and no check is performed if the abbreviations are compatible with other functions.

Value

If Return='Matrix', an N x N square matrix, with N equal to the number of rows in Pedigree (after running PedPolish) or the number of unique individuals in Pairs. If Return='Array', an N x N x R array is returned, with R, the number of different relationships, determined by GenBack and patmat.

The following abbreviations are used within the returned Matrix, or as names of the 3rd dimension in the Array or of the List:

`S`	Self
`M`	Mother
`P`	Father
`MP`	Mother or Father (`patmat=FALSE`)
`O`	Offspring
`FS`	Full sibling
`MHS`	Maternal half-sibling
`PHS`	Paternal half-sibling
`XHS`	other half-sibling (hermaphrodites)
`HS`	half-sibling (`patmat=FALSE`)
`MGM`	Maternal grandmother
`MGF`	Maternal grandfather
`PGM`	Paternal grandmother
`PGF`	Paternal grandfather
`GP`	Grandparent (`patmat=FALSE`)
`GO`	Grand-offspring
`FA`	Full avuncular; maternal or paternal aunt or uncle.
`FN`	Full nephew/niece
`HA`	Half avuncular
`HN`	Half nephew/niece
`DFC1`	Double full first cousin
`FC1`	Full first cousin
`U`	Unrelated (or otherwise related)

Examples

Rel.griffin <- GetRelM(Ped_griffin, patmat=TRUE, GenBack=2)
table(as.vector(Rel.griffin))
# turning matrix into vector first makes table() much faster
PlotRelPairs(Rel.griffin)

sequoia documentation built on Sept. 8, 2023, 5:29 p.m.