yunruilu / haps

1. Installation locally

Please download the compressed R package here: HAPS

Or download the compressed R package and readme file together here:HAPS+Readme file

1.1 On Windows system

Make sure that your R is installed in 'c:\program files'

Install Rtools in 'c:\program files'

Add R and Rtools to the Path Variable on the Environment Variables panel, including

c:\program files\Rtools\bin

c:\program files\Rtools\gcc-4.6.3\bin

c:\program files\R\R.3.x.x\bin\i386

c:\program files\R\R.3.x.x\bin\x64

Run the following code in R

install.packages('HAPS_0.1.0.tar.gz', repos = NULL, type='source')

1.2 On Linux and Mac systems

Just run the following code in R

install.packages('HAPS_0.1.0.tar.gz', repos = NULL, type='source')

2. Basic example

There are three functions in HAPS.

• haplo.GWAS perform the GWAS using the haplotype data

haplo.GWAS(Strfile=NULL,Manhattan=TRUE,Genefile=system.file("examples/genotype.sample.txt.gz", package="HAPS"),Phefile = system.file("examples/phenotype.sample.csv", package="HAPS"),Chrom ="all",max.merge=5,num.comp=3,p.adjust.method ="fdr",threshold=0.05)

• haplo.GS perform the genomic prediction using the haplotype data

haplo.GS(Strfile=NULL,Genefile1=system.file("examples/training.geno.txt.gz",package="HAPS"),Phefile1=system.file("examples/training.phe.csv",package="HAPS"),Genefile2=system.file("examples/predict.geno.txt.gz", package="HAPS"),max.merge=3,num.comp=3)

• haplo.CV.GS perform the genomic prediction using the haplotype data through the cross validation

haplo.CV.GS(Strfile=NULL,Genefile=system.file("examples/training.geno.txt.gz", package="HAPS"),Phefile = system.file("examples/training.phe.csv", package="HAPS"),Chrom ="all",max.merge=3,num.comp=3,nfold=10)

3. parameter setting

3.1 haplo.GWAS

Strfile: file name of the population structure. The first column of this file is the line, the other columns are the population structure data. If the file dose not exist, Strfile=NULL

Manhattan: a logical value ( TRUE/FALSE) whether output manhattan plots

Genefile: the file name of genotype data

Phefile: the file name of phenotype data

Chrom: a number represents which chromsome is analyzed. 1 for the first chromosome; "all" for the whole chromosome

max.merge: an integer. When the number of haplotypes is less than the max.merge, then these haplotypes are combined into one type haplotype, default is 3, 1 means all the haplotype do not be combined. The higher the number, the faster the computation speed. This parameter depends on your population size, please adjusting the parameter to get the perform results

num.comp: a number indicating how many components are selected as the fixed effects, default is 3

p.adjust.method: method for adjust p value, including "holm", "hochberg", "hommel", "bonferroni", "BH", "BY", "fdr", "none"

threshold: a threshold value for significant tests, 0.05/0.01/defind any number

3.2 haplo.GS

Strfile: file name of the population structure The first column of this file is the line, the other columns are the population structure data. If the file dose not exist, Strfile=NULL

Genefile1: the file name of genotype data for training population

Phefile1: the file of phenotype data for training population

Genefile2: the file name of genotype data for predicted population

max.merge: the same to the parameter in haplo.GWAS

num.comp: the same to the parameter in haplo.GWAS

3.3 haplo.CV.GS

Strfile: the file name of genotype data

Genefile: the file name of genotype data

Phefile: the file name of phenotype data

Chrom: the same to the parameter in haplo.GWAS

max.merge: the same to the parameter in haplo.GWAS

num.comp: the same to the parameter in haplo.GWAS

nfold: a number indicating how many folds were used to perform the cross validation