QTL mapping for function-valued traits using estimating equations.
Xiong H, Goulding EH, Carlson EJ, Tecott LH, McCulloch CE, Sen S (2011) "A flexible estimating equations approach for mapping function-valued traits" Genetics 189:305-316.
Install the most current stable version using the following commands
which will install the
stable branch; this should run on all
install.packages("devtools") library(devtools) install_bitbucket("linen/qtlcurve",ref="stable")
default branch contains code under development. You likely don't
want this branch, but if you do, use:
withcpp branch of the package contains C++ code that makes the
functions a lot faster. Currently, that feature is in development,
and compiles reliably on Linux only. To get that branch use:
Read in the example data from Xiong et. al. (2011).
# get name of genotype/phenotype and phenotype covariate files # from system location genofile <- system.file("extdata","xiong11.csv",package="qtlCurve") pcovfile <- system.file("extdata","xiong11phenocov.csv", package="qtlCurve") # read in data and assign as cross object cr <- readCross(genoPhenoFile=genofile,phenoCovFile=pcovfile)
First we have to calculate the genotype probabilities for Haley-Knott regression.
# calculate genotype probabilities for haley-knott regression cr <- calc.genoprob(cr,step=1) # calculate one-dimensional genome scan # use b-spline basis matrix of dimension 6 # select out the activity data columns out1 <- scanOne(cr,pheno.col=3:224, basisMatrix=basisMatrix("bs",x=1:222,dim=6)) # plot genome scan plot(out1)