geneSetAnalysis: Pathway and Gene Set Analysis using Density Curves

Description Usage Arguments Details Value Author(s) See Also

Description

Look for differentially expressed (DE) Pathways and Gene Sets from differential gene expression results.

Usage

1
2
3
4
5
6
7
8
geneSetDensity(deList, geneSets, speciesID = getCurrentSpecies(), colorset = c(2:(length(deList) + 1)), 
	optionsFile = "Options.txt", results.path = NULL, folderName = "", 
	toolName = c("MetaResults", "RoundRobin", "RankProduct", "SAM", "DESeq", "EdgeR"), 
	descriptor = "GeneSet", minGenesPerSet = 2, trimGenesToGeneMap = TRUE,
	geneMapColumn = if (speciesID 
	cutPvalue = 0.05, cutRankShift = NULL, cutFold = 0.1, cutFDR = 0.05,
	makePlots = TRUE, PLOT.FUN = NULL, makeGeneTables = TRUE,
	doFDR = TRUE, NgeneSets = 500)

Arguments

deList

a named list of differential expression (DE) data frames (2 or more), containing the DE results for each condition. The names of the list should be the names of the conditions.

geneSets

either a character vector of names of pre-defined GeneSet objects, or a list of vectors of GeneIDs.

speciesID

the species (organism) being evaluated

colorset

a vector of colors for displaying each condition

results.path

top level folder of results, defaults to the value in the Options table

folderName

the name of the subfolder of DE comparisons to use

toolName

the name of the DE tool used to generate the differential gene expression results.

descriptor

a character string for naming the final folder of created analyses.

minGenesPerSet

the minimum number of genes that a GeneSet can contain. Very low count sets are ignored.

trimGenesToGeneMap

logical, removes extra gene terms from any gene set references that are not in the current gene map.

geneMapColumn

the name of the column in the current species' GeneMap that contains the GeneIDs found in the GeneSets. Some species use full notation GeneIDs and others uses simpler gene symbols.

cutPvalue

maximum P-value to include a GeneSet in the final results. Lowering the P-value will produce less results.

cutRankShift

minimum rank shift to include a GeneSet in the final results. Raising the ranks shift will produce less results.

cutFold

minimum fold change to include a GeneSet in the final results. Raising the fold change will produce less results.

cutFDR

maximum FDR to include a GeneSet in the final results. Lowering the FDR will produce less results. Only used when doFDR=TRUE.

doFDR

logical, should a False Discovery Rate be calculated, and added to all results. Rather slow.

makePlots

logical, should plot images for each GeneSet be created

PLOT.FUN

optional function to be called for making gene set plots. Must take one argument that is a vector of gene IDs.

makeGeneTables

logical, should small HTML files of gene-to-GeneSet membership tables be created

Details

This function analyzes how pre-defined sets of genes are differentially expressed between 2 or more conditions. The methodology is to treat each gene as a point mass, located at its rank order location in each of the condition's DE results, and then summarize all the genes in each GeneSet as a density mass for the set. A gene set is called differentially expressed in a condition if its density mass is significantly shifted with respect to that same gene set in all other conditions.

Value

A folder of results, where the final name/location is a combination of results.path, toolName, speciesID, folderName, descriptor.

In that folder will be a family of HTML files for each condition, one HTML and .txt file ocvering all GeneSets and conditions, and a subfolder of all plots and hyperlinked gene/module files.

Briefly, for each condition, there are HTML files for the most up-regulated and down-regulated GeneSets, with P-values and density rank shifts for each significantly DE GeneSet. Within each HTML file are links to images of density plots for each GeneSet, and to gene membership tables to examine the details about genes in that GeneSet.

Author(s)

Bob Morrison

See Also

geneSetEnrichment for a similar function based on enrichment instead of mass density. pipe.GeneSetAnalysis for a high-lever wrapper function using SampleIDs.


robertdouglasmorrison/DuffyTools documentation built on Dec. 7, 2018, 8:02 a.m.