racExSet-class {GGtools}R Documentation

Class "racExSet" for combining RareAlleleCount representations of SNPs, gene expression data, and other phenotype data

Description

A container for combining RareAlleleCount representations of SNPs, gene expression data, and other phenotype data

Objects from the Class

Objects can be created by calls of the form new("racExSet", phenoData, experimentData, annotation, exprs, racs, ...).

Slots

racAssays:
Object of class "AssayData" SxN matrix of 0, 1, 2, or NA, rownames are RS numbers from dbSNP, column names are sampleNames
assayData:
Object of class "AssayData" GxN matrix of expression values
rarebase:
character vector of length G stating the rare base of the SNP
SNPalleles:
character vector of length G stating the alleles in the form "A/B"
phenoData:
Object of class "AnnotatedDataFrame" standard phenoData instance
experimentData:
Object of class "MIAME" experiment level metadata
annotation:
Object of class "character" tag
.__classVersion__:
Object of class "Versions" R class version control metadata

Extends

Class "eSet", directly. Class "VersionedBiobase", by class "eSet". Class "Versioned", by class "eSet".

Methods

snps
signature(x = "racExSet"): accessor for the racAssay matrix
snpNames
signature(x = "racExSet"): give RS numbers for rows of racAssay matrix
racAssays
signature(x = "racExSet"): accessor for the racAssays slot
oneFit
signature(racExSet = "racExSet", geneid = "genesym", snpid = "character", fitfun = "function"): use a genesym instance (e.g., genesym("DDR1") to identify the expression values, an rs number to identify the snp genotype information, and fitfun to fit a model with expression as dependent variable and genotype (rare allele counts) as predictor. Any fitting function accepting a formula and data frame should succeed (e.g., lm).
oneFit
signature(racExSet = "racExSet", geneid = "character", snpid = "character", fitfun = "function"): as above, but can avoid the lookup of the gene symbol by supplying the probe set identifier directly. This is problematic at present as most probe set identifier strings are not valid R variable names, and so cannot serve directly in the formula. Some backtick coding is needed to get this to work generally.
oneFit
signature(racExSet = "racExSet", geneid = "list", snpid = "character", fitfun = "function"): the geneid parameter must be a named list of length one with content a numeric vector of expression values conforming to the snp values. Avoids lookup of gene name and of expression values.
oneFit
signature(racExSet = "racExSet", geneid = "list" , snpid="formula", fitfun="function"): the geneid parameter must be bound to a named list of length 1 with content a numeric vector with one element per sample; the name of the list element is the gene symbol on which the numeric vector is a vector of expression values; the snpid parameter is bound to a formula template, either ~. or ~factor(.) expressing choice between an additive genetic model and a general model.

Notes

Helper functions SNPalleles and rarebase are accessors. snps5PrimeTo(gene,rad) will obtain rs numbers for SNPs in selected locations relative to the specified gene. snps3PrimeTo and snpsNear have similar purposes.

Note

The genesym class works in conjunction with annotation to allow translation from HUGO symbols to affy probe set names according to Bioconductor annotation environments. The getpsid function carries out the translation; the genesym class/wrapper just tells us that a string will be serving the role of a HUGO symbol and allows us to dispatch accordingly. The snpID class has similar role. It just lets us know that a string is intended to be regarded as an rs number.

hgfocusSYMMEMO is a lookup table quickly establishing the probeset ID based on a HUGO symbol; likewise mgu74av2SYMMEMO.

chr20GGdem is a demonstration example of chromosome 20 hapmap snps and hgfocus expression data from GEO as generated by Cheung and colleagues.

gse2031GG is a demonstration example of BXD mouse snps and mgu74av2 expression data from GSE2031 in GEO.

The exclMono function eliminates loci that happen to be monomorphic in the sample.

Author(s)

VJ Carey <stvjc@channing.harvard.edu>

See Also

make_racExSet for construction; eSet ancestor class

Examples

example(make_racExSet)
dem
snpNames(dem)[1:4]
featureNames(dem)[1:4]
oneFit(dem, genesym("DDR1"), "rs4814683", lm)
ddr1 = list(DDR1=exprs(dem)["1007_s_at",])
oneFit(dem, ddr1, "rs4814683", aov)
data(chr20GGdem)
ddr1 = list(DDR1=exprs(chr20GGdem)["1007_s_at",])
summary(oneFit(chr20GGdem, ddr1, ~rs4814683+factor(rs6060535), lm))
dim(snps(chr20GGdem))
dim(snps(exclMono(chr20GGdem)))

[Package GGtools version 1.4.0 Index]