diffPathways: Detect differentially activated pathways between fingerprints

In pathprint: Pathway fingerprinting for analysis of gene expression arrays

Description

A function to return pathways consistently differentially expressed between two groups of pathway fingerprints

Usage

diffPathways(fingerprints, fac, threshold)

Arguments

fingerprints	matrix of fingerprints, the number of columns should correspond to the length of fac
fac	vector of characters or factors, in an order corresponding to the order of columns in the fingerprint matrix. Contains two levels, denoting the groups to be compared.
threshold	numeric, between 0 and 2 - the threshold at which to assign an average difference in pathway usage.

Details

The vector of factors must contain only two levels (or two unique values for a character vector).

Value

Returns a list of the rownames (i.e. pathways for the pathway fingerprint) corresponding to the rows for which the difference in the means between the two groups is greater than the threshold value.
For a ternary fingerprint (-1,0,1), setting the threshold between 0.5 and 1 ensures that rownames are selected that differ across the majority of the arrays in the two groups. with values closer to 1 representing higher stringency. This can break down and allow false positives in the case where one group contains a significant but minority number of +1 and the other -1s.

Author(s)

Gabriel Altschuler

References

Altschuler, G. M., O. Hofmann, I. Kalatskaya, R. Payne, S. J. Ho Sui, U. Saxena, A. V. Krivtsov, S. A. Armstrong, T. Cai, L. Stein and W. A. Hide (2013). "Pathprinting: An integrative approach to understand the functional basis of disease." Genome Med 5(7): 68.

See Also

exprs2fingerprint, consensusDistance, consensusFingerprint

Examples

require(pathprintGEOData)

# Use ALL dataset as an example

require(ALL)
data(ALL)
annotation(ALL)
library(SummarizedExperiment)

# load  the data
data(SummarizedExperimentGEO)

ds = c("chipframe", "genesets","pathprint.Hs.gs",
    "platform.thresholds","pluripotents.frame")
data(list = ds)

# extract part of the GEO.fingerprint.matrix and GEO.metadata.matrix
GEO.fingerprint.matrix = assays(geo_sum_data[,300000:350000])$fingerprint
GEO.metadata.matrix = colData(geo_sum_data[,300000:350000])

# free up space by removing the geo_sum_data object
remove(geo_sum_data)

# The chip used was the Affymetrix Human Genome U95 Version 2 Array
# The correspending GEO ID is GPL8300

# Analyze patients with ALL1/AF4 and BCR/ABL translocations
ALL.eset <- ALL[, ALL$mol.biol %in% c("BCR/ABL", "ALL1/AF4")]
ALL.exprs<-exprs(ALL.eset)

patient.type<-as.character(ALL$mol.biol[
    ALL$mol.biol %in% c("BCR/ABL", "ALL1/AF4")])

# Process fingerprints
ALL.fingerprint<-exprs2fingerprint(exprs = ALL.exprs,
    platform = "GPL8300",
    species = "human",
    progressBar = TRUE
)

color.map <- function(mol.biol) {
    if (mol.biol=="ALL1/AF4") "#00FF00" else "#FF00FF"
    }
patientcolors <- sapply(ALL$mol.biol[
    ALL$mol.biol %in% c("BCR/ABL", "ALL1/AF4")],
    function(x){
    if (x == "ALL1/AF4") "#00FF00" else "#FF00FF"
    })


# define differentially activated pathways between the two groups
signif.pathways<-diffPathways(ALL.fingerprint,
fac = patient.type,
threshold = 0.6)

# draw heatmap
heatmap(ALL.fingerprint[signif.pathways,],
    ColSideColors = patientcolors,
    col = c("blue", "white", "red"),
    scale = "none", mar = c(10,20),
    cexRow = 0.75)
title(sub = "Pathways differentially activated in patients
with ALL1/AF4 (green) and BCR/ABL(purple) translocations",
        cex.sub = 0.75)

pathprint documentation built on April 28, 2020, 7:54 p.m.