snpMatrixScour: SNP filtering based on Minor Allele Frequency, Hardy-Weinberg...
In GeneGeneInteR: Tools for Testing Gene-Gene Interaction at the Gene Level

Description Usage Arguments Details Value See Also Examples

snpMatrixScour aims at filtering out SNPs of a snpMatrix object based on Minor Allele Frequency criterion, deviation to Hardy-Weinberg Equilibrium and SNPs call rate.

1 2	snpMatrixScour(snpX, genes.length = NULL, genes.info = NULL, min.maf = 0.01, min.eq = 0.01, call.rate = 0.9)

`snpX`	SnpMatrix object from which SNPs are to be removed
`genes.length`	(optional) numeric vector. It is the length (in columns/SNPs) of each gene. Each gene declared is considered contiguous with the one before and after it. `genes.lengths` can be named (names will be kept).
`genes.info`	(optional) a `data.frame` with four columns named `Genenames`, `SNPnames`, `Position` and `Chromosome`. Each row describes a SNP and missing values are not allowed.
`min.maf`	a single value between 0 and 0.5 that gives the threshold for the MAF (Minor Allele Frequency) of a SNP. SNP with MAF < `min.maf` are removed. Default is 0.01.
`min.eq`	a single value between 0 and 1 that gives the maximum acceptable p-value for the χ ^2 verifying HWE deviation. SNP that does not meet that criterion are removed. Default is 0.01.
`call.rate`	a single value between 0 and 1 that gives the minimum acceptable call rate for a SNP. Default is 0.9. Low values for SNPs call rate can make imputation harder (residual missing values).

This function removes SNPs that does not meet all following criteria:

MAF < min.maf,
HWE < min.eq, where HWE is the p-value of the χ ^2 test of deviation to Hardy-Weinberg Equilibrium,
Call rate > call.rate.

If genes.length and genes.info are provided by the user, an updated version is returned by snpMatrixScour. The returned object can be directly used as inputs of the GGI function.

A list with two objects:

snpX: the SnpMatrix object where non-conform SNPs are removed.
genes.info: the object that contains the updated gene lengths information. Can be a numeric vector (possibly named) or a data frame. If genes.length and genes.info are not provided by the user as input of the snpMatrixScour function, the genes.info object is NULL.

GGI

## Not run: 
ped <- system.file("extdata/example.ped", package="GeneGeneInteR")
info <- system.file("extdata/example.info", package="GeneGeneInteR")
posi <- system.file("extdata/example.txt", package="GeneGeneInteR")
data.imported <- importFile(file=ped, snps=info, pos=posi, pos.sep="\t")

## End(Not run)
### Equivalent loading of the imported data
load(system.file("extdata/dataImported.Rdata", package="GeneGeneInteR"))

## In this example, SNPs are with MAF lower than 0.2 or p-value for HWE testing lower than 0.05 or
# a proportion of missing value higher than 0.2 are removed
data.scour1 <- snpMatrixScour(data.imported$snpX, genes.info = data.imported$genes.info,
                               min.maf = 0.2, min.eq=0.05, call.rate = 0.8)
## Two genes have been completely removed from the resulting dataset.