mvGST (version 1.0.0)

p.adjust.SFL: Short Focus Level adjustment

Description

Takes a named numeric vector of raw p-values as input and returns the Short Focus Level adjusted p-values, where the adjustment is based on controlling the FWER at a specified level within the structure of the GO graph.

Usage

p.adjust.SFL(rawp, ontology=c("BP","CC","MF"),  focus='rn', ancestors, offspring, trace=FALSE,  recycle=TRUE, sig.level=0.05)

Arguments

rawp

named numeric vector of p-values where the names correspond to the GO ID for which the provided p-values correspond to the given GO Term. These must be 'two-sided' p-values, i.e., from a two-sided test.

ontology

The ontology of interest. Must be one of 'BP', 'CC', or 'MF'. All names of rawp must be from the same ontology. Defaults to 'BP'.

focus

the focus level of interest. Default is set at the root node, the logical place to start if there is no better place to start.

ancestors

named lists corresponding to the ancestor and offspring structure of all named GO IDs in rawp. These are optional, and simply result in faster computation of the adjusted p-values if they are already available from the R session when p.adjust.SFL is called.

offspring

named lists corresponding to the ancestor and offspring structure of all named GO IDs in rawp. These are optional, and simply result in faster computation of the adjusted p-values if they are already available from the R session when p.adjust.SFL is called.

trace

logical denoting whether or not progress about the algorithm is output to the user. Defaults to FALSE.

recycle

logical determining whether or not to recycle any threshold corresponding to a rejected leaf node back into the GO graph. Defaults to TRUE, as it can result in greater power for the Short Focus Level method at a slight increase to the computational burden.

sig.level

numeric value at which to control the family-wise error rate within the structure of the GO graph.

Value

returns the adjusted p-values with naming and ordering identical to the original "rawp" values.

Details

To access the tutorial document for this package, type in R: vignette("mvGST")

References

Saunders G., Stevens J.R., and Isom S.C. "A shortcut for multiple testing on the directed acyclic graph of Gene Ontology." BMC Bioinformatics 2014 (under review).

Saunders, G., 2014. "Family-wise error rate control in QTL mapping and gene ontology graphs with remarks on family selection." PhD thesis, Utah State University, Department of Mathematics and Statistics. http://digitalcommons.usu.edu/etd/2164/

Examples

# Get GO terms of interest
library(GOstats); library(annotate)
GO.vec <- c("GO:0001775","GO:0007275")
g <- GOGraph(GO.vec, GOBPPARENTS)
g <- removeNode("all",g)
GOids <- names(nodes(g))

# Get p-values for all GO terms of interest
# (here, simulated for demonstration)
# Make sure names are GO term IDs
set.seed(1)
rawp <- rbeta(length(GOids), .2, 1)
names(rawp) <- GOids

# P-value adjustment using Short Focus Level
padj <- p.adjust.SFL(rawp, ontology='BP')
head(padj)
# These are in the same order as rawp, with
# names corresponding to GO terms.
# Calling GO terms significant when padj is 
# less than alpha controls the FWER at alpha,
# within the context of the GO graph.

# See package vignette for larger examples with discussion: 
#    vignette("mvGST")