[BioC] gseattperm return all NA

Saroj Mohapatra smohapat at vbi.vt.edu
Thu Oct 16 00:53:20 CEST 2008


Hi all:

I have a situation in which I am not able to interpret the output of the 
function gseattperm in the package Category.

I have six arrays belonging to two groups. I am able to get some 
pathways that are different between the two groups using the KEGG-based 
GSEA method (described in section 13.2.2 of the chapter 13 (Gene Set 
Enrichment Analysis) of the book Bioconductor Case Studies (Hahne et al 
2008). The KEGGmnplot and heatmap are similar to what is described in 
that section.

However when I try permutation testing (section 13.2.3), gseattperm 
returns a matrix of two columns with all missing values. I am not sure 
if it is unusual.

I copy the code below. Thanks for any feedback.

Saroj

-----------------------------------
# Non-specifice filtering to reduce low-varibility genes:
 > mydat.f = nsFilter(mydat, remove.dupEntrez = FALSE, var.cutoff = 
0.5)$eset

# Get the collection of genesets for the data and compute incidence matrix:
 > gsc = GeneSetCollection(mydat.f, setType = KEGGCollection())
 > Am = incidence(gsc)
 > (dim(Am))
[1] 195 5119

# Retain only the genes mentioned in the incidence matrix:
 > nsF = mydat.f[colnames(Am), ]
 > nsF$Treatment = as.factor(nsF$Treatment)

# Reduce the incidence matrix by removing all gene sets that have fewer 
than ten genes in them:
 > selectedRows = (rowSums(Am) > 10)
 > Am2 = Am[selectedRows, ]
 > dim(Am2)
[1] 173 5119

# Run permutation testing:
 > NPERM = 1000
 > pvals = gseattperm(nsF, nsF$Treatment, Am2, NPERM)
 > (sum(is.na(pvals[, 1])))
[1] 173
 > (sum(is.na(pvals[, 2])))
[1] 173


 > sessionInfo()
R version 2.7.2 (2008-08-25)
x86_64-unknown-linux-gnu

locale:
LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=C;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C

attached base packages:
[1] splines tools stats graphics grDevices utils datasets
[8] methods base

other attached packages:
[1] mouse4302.db_2.2.0 GSEABase_1.2.2 Category_2.6.0
[4] annotate_1.18.0 xtable_1.5-3 AnnotationDbi_1.2.2
[7] RSQLite_0.7-0 DBI_0.2-4 graph_1.19.6
[10] genefilter_1.20.0 survival_2.34-1 Biobase_2.0.1

loaded via a namespace (and not attached):
[1] cluster_1.11.11 RBGL_1.17.2 XML_1.96-0
-----------------------------------------------


More information about the Bioconductor mailing list