rJPSGCS (version 0.2-10)

GeneDrops: simulate genotypes on a pedigree

Description

This function is an R wrapper for the JPSGCS multi-locus gene-drop program.

Usage

GeneDrops(ld.par = "ld.par", ped = "in.ped", n, complete.data=FALSE,
gd.ped = "gd.ped", compress.pedfiles=FALSE)

Arguments

ld.par

Name of a LINKAGE parameter file or an LD model file output by FitGMLD.

ped

Name of the post-MAKEPED pedfile (see the explanation in the documentation for read.pedfile) giving the pedigree relationships and marker data availability to simulate.

n

Number of pedigrees to simulate.

complete.data

Should complete data be simulated; i.e., should we ignore the marker data availability in ped? Default is FALSE.

gd.ped

Name of the output file of simulated marker data in post-MAKEPED pedigree file format.

compress.pedfiles

Should the output pedigree file be compressed? Default is FALSE.

Value

No value. The output is in the file named by gd.ped.

See Also

FitGMLD

Examples

# NOT RUN {
# 1. Gene-drop assuming independent loci in the founders, using allele 
# frequencies and map information from the example data set. 
data(exdat)
# Print out the data to examine missing data pattern.
exdat$markers
#Coerce to snp.matrix
sdat<-as(exdat$markers,"snp.matrix") 
# Need LINKAGE parameter and pedigree files as input.
# For independent loci in the founders use a standard LINKAGE parameter file.
write.parfile(snp.data=sdat, map=exdat$map, file="indep.par")
# To construct the pedigree file for unrelated individuals, use
# the pedinf="unrelated" option.
write.pedfile(pedinf="unrelated",snp.data=sdat,file="unrel.ped")
GeneDrops(ld.par="indep.par",ped="unrel.ped",n=10,gd.ped="gd.ped")
# Read the SNP data back into R. 
simdat<-read.pedfile("gd.ped")$snp.data
# Look at the simulated data: need to coerce to numeric matrix to see
# values of individual genotypes, rather than the usual snp.matrix summary
as(simdat,"numeric")

# 2. Simulate data at 5 independent SNPs 0.1 cM apart on a parent-child trio.
map=(1:5)/10
# Marker data on each individual as follows:
mom=c(1,2,1,1,1) # complete data
dad=c(0,NA,0,1,0) # missing data at SNP 2
child=c(NA,1,1,2,NA) #missing data at SNPs 1 and 5
sdat<-rbind(mom,dad,child)
colnames(sdat)<-paste("SNP",1:5,sep="")
sdat<-as(sdat,"snp.matrix")
write.parfile(snp.data=sdat, map=map, file="trio.par")
# Block of code setting up pedigree information for mom, dad and child.
# See the Details section of the write.pedfile help file for 
# information on the file format.
trioinf<-matrix(c(
    1, 1, 0, 0, 0, 0, 0, 2, 1, #mom: ID=1, no parents in pedigree
    1, 2, 0, 0, 0, 0, 0, 1, 1, #dad: ID=2, no parents in pedigree
    1, 3, 2, 1, 0, 0, 0, 2, 1),#child: ID=3, dadID=2, momID=1
    byrow=TRUE, ncol=9, nrow=3)
write.pedfile(pedinf=trioinf, snp.data=sdat,file="trio.ped")
GeneDrops(ld.par="trio.par",ped="trio.ped",n=10,gd.ped="gd.ped")
#Read the SNP data back into R. 
simdat<-read.pedfile("gd.ped")$snp.data
# Look at the missing data patterns in simdat
as(simdat,"numeric")

# }
# NOT RUN {
# 3. Repeat gene drop for the trio, but this time use an LD model fit 
# by FitGMLD. FitGMLD requires a LINKAGE parameter file and a transposed
# pedigree file as input. Can use the parameter file trio.par again.
# The transposed pedigree file can be generated as follows:
write.pedfile(pedinf=trioinf,snp.data=sdat,file="f.ped",transpose=TRUE)
FitGMLD("trio.par","f.ped","out.ld.par")
#Call to GeneDrops uses the output of FitGMLD as the parameter file and
# the regular (not transposed) pedigree file 
GeneDrops(ld.par="out.ld.par",ped="trio.ped",n=10,gd.ped="gd.ped")
simdat<-read.pedfile("gd.ped")$snp.data
unlink(c("f.ped","out.ld.par"))
# }
# NOT RUN {
unlink(c("unrel.ped","trio.ped","indep.par","trio.par", "gd.ped"))
# }