The basis of genetic likeness

Overview

We will learn how to compute an additive relationship matrix from pedigree using the tabular method.

Reading file

Use the function read_excel in the readxl package to read the pedigree file (Simdata.xlsx) in a data frame format.

install.packages("readxl")
library(readxl)
dat <- read_excel(file.choose())
dim(dat)
head(dat)

synbreed package

The synbreed package includes the function kin for computing a relationship matrix. Learn more about the Synbreed project. The kin function requires the object of class gpData for inputs.

install.packages("synbreed")
library(synbreed)
gp.ped <- create.pedigree(ID=dat$Animal, Par1=dat$Sire, Par2=dat$Dam, unknown = "0")
head(gp.ped)
gpData<- create.gpData(pedigree = gp.ped)
class(gpData)

We can obtain an additive relationship matrix by setting the argument ret equals to add.

A1 <- kin(gpData, ret = "add")
dim(A1)
table(colnames(A1) == rownames(A1))

One thing we need to be cautious of is that the kin function does not retrain the original order.

head(colnames(A1))
head(dat$Animal)

This brings up a problem when we want to link phenotypic values to the A matrix because animals appear in a different order. So let’s keep the original order of animal IDs in the dat object.

index <- match(dat$Animal, rownames(A1))
A <- A1[index, index]
dim(A)

The A object is a reordered A matrix where animals appear in the same order as that of the dat object.

head(rownames(A))
head(dat$Animal)
table(dat$Animal == rownames(A))
table(dat$Animal == colnames(A))

kinship2 package

Alternatively, we can use the kinship2 package to compute a relationship matrix. The kinship function returns the coefficient of kinship.

install.packages("kinship2")
library(kinship2)
A2 <- kinship(dat$Animal, dat$Sire, dat$Dam) 
dim(A2)
table(colnames(A2) == rownames(A2))
table(dat$Animal == rownames(A2))

Save R object

Save the object A so that we can reuse in the next class.

save(A, file = "A.Rda")