getMatMI: Computation of the mutual information between SNP pairs in...

In MathieuEmily/EpiTag: TagSNP selection for epistasis

Description

getMatMI allows the computation of the mutual information between SNP pairs within 1 genomic regions or between 2 genomic regions.

Usage

getMatMI(Region1,Region2=NULL,allelic=FALSE)

Arguments

Region1	A nxp1 matrix or data.frame containing allele or genotype of the p1 SNPs for the n individuals.
Region2	A nxp2 matrix or data.frame containing allele or genotype of the p2 SNPs for the n individuals. Region 1 and Region 2 must have the same number of rows.
allelic	A flag indicating whether data are allelic allelic=TRUE or genotypic allelic=FALSE.

Details

getMatMI is an implementation of the statistical procedure proposed in Friguet and Emily (2013). When considering only 1 genomic region (i.e. when Region2=NULL), getMatMI computes the pairwise mutual information between all pairs of SNPs : MI((i,j),(k,l)) where MI is the Mutual Information, (i,j) two SNPs within the genomic region and (k,l) two SNPs within the same genomic regions. When considering only 2 genomic regions (i.e. when Region2!=NULL), getMatMI computes the pairwise mutual information between pairs of SNPs across the two regions: MI((i,j),(k,l)) where MI is the Mutual Information, (i,j) two SNPs within the first genomic region and (k,l) two SNPs within the second genomic regions.

Value

An object of class MIMatrix which is a list containing the following components:

mat.MI	A a p1xp2xp1xp2 array that stores the mutual information between all pairs of SNPs
nb.locus	An integer with two possible value: nb.locus=1 corresponding to the computation of the mutual information within 1 genomic region (i.e. when Region2=NULL) and nb.locus=2 for the computation of the mutual information between 2 genomic regions (i.e. when Region2!=NULL).

References

Friguet, C. and Emily, M. Selection de marqueurs pour la detection d'interactions de genes. Proceedings of 45emes Journees de Statistique, May 2013, Toulouse, France. pp.178, 2013.

See Also

EpiTag

Examples

############
## Example with simulated allelic data
############
data(RegionA)
data(RegionB)
MatMI <- getMatMI(Region1=RegionA,Region2=RegionB,allelic=TRUE)

############
## Example with a real subset of genotypic data
############
## Loading of the genotypic data
load(system.file("extdata/Chr15_20S.Rdata", package="EpiTag"))
## Computation of the mutual information
MatMI15 <- getMatMI(Region1=Chr15.20S)

MathieuEmily/EpiTag documentation built on May 8, 2019, 9:48 a.m.