[BioC] globaltest

Caroline Reiff C.Reiff at rri.sari.ac.uk
Tue Mar 20 15:37:09 CET 2007


Hi,
I am trying to use the package "globaltest" on samples with 2 
treatments and could do with some help.....
For input "examplesX" I have arranged my normalised data in a matrix 
with rows=probe-ids, columns=samples
For input "examplesY" I have a vector with the data description (WT or MT)
For the genesets I have a list containing probeids for 2 pathways

My problem: I don`t seem to be able to read my probeids from a .txt 
file correctly into a vector called "pathway1"

This is how I tried to do it in bioconductor:

examplesX<-as.matrix(read.table("normalisedresulttable6reps.txt",header=TRUE)) 

examplesY<-c("WT","MT","WT","MT","WT","MT")
pathway1<-as.vector(read.table("Immune-associated-probesetids.txt"))
pathway2 <- as.vector(read.table("Inflammatory 
response-associated-probesetids.txt"))
lst <- list(Probes.immune=(pathway1),Probes.Inflammation=(pathway2))
 > lst
$Probes.immune
                      V1
1            1416292_at
2          1426044_a_at
3            1426207_at
4            1440085_at


$Probes.Inflammation
                      V1
1            1425947_at
2            1460251_at
3          1416381_a_at
4            1416657_at

//test all genes, this works..but not sure whether I have to specify 
that there are 2 treatments?
gt <- globaltest(examplesX, examplesY,)
gt

//test only those genes in the genelists in lst, this does not work
gt <- globaltest(examplesX, examplesY,lst)
gt
Error: genesets should be a (list of) vector(s)

If I however make my list lst as follows my "pathway1" "pathway2" and 
"lst" looks different and globaltest works:
 > 
pathway1<-c("1416292_at","1426044_a_at","1426207_at","1440085_at","NuGO_emt025263_at","NuGO_emt064290_at","1415697_at","1418133_at")
 > 
pathway2<-c("1416292_at","1426044_a_at","1426207_at","1440085_at","NuGO_emt025263_at","NuGO_emt064290_at","1415697_at","1418133_at")
 > lst <- list(Probes.immune=(pathway1),Probes.Inflammation=(pathway2))
 > lst

$Probes.immune
[1] "1416292_at"        "1426044_a_at"      "1426207_at"
[4] "1440085_at"        "NuGO_emt025263_at" "NuGO_emt064290_at"
[7] "1415697_at"        "1418133_at"

$Probes.Inflammation
[1] "1416292_at"        "1426044_a_at"      "1426207_at"
[4] "1440085_at"        "NuGO_emt025263_at" "NuGO_emt064290_at"
[7] "1415697_at"        "1418133_at"

 > gt <- globaltest(examplesX, examplesY,lst)
 > gt

Global Test result:
Data: 6 samples with 23865 genes; 2 pathways tested
Model: logistic
Method: All 10 permutations

                     Genes Tested Statistic Q Expected Q sd of Q
Probes.immune           8      8       80.72      35.02  39.695
Probes.Inflammation     8      8       80.72      35.02  39.695
                     P-value
Probes.immune           0.2
Probes.Inflammation     0.2



More information about the Bioconductor mailing list