Problem with cummeRbund diffData()

[Enrique Zudaire]

Hi all,

I'm running Tophat/cufflinks/cuffdiff for differential gene expression and analysis with cummeRbund (v 2.0.0). I'm having an issue with diffData(). I have restricted significance results to one pairwise comparison with getSig (). This works great. However, when I try to do the same with getDiff() comparisons between all samples are reported. I have included the code below. Any thoughts on why this does not work?

...

diffGenesIds<-getSig(cuff,x='B73',y='B74',alpha=0.05,level='genes')
diffGenes<-getGenes(cuff,diffGenesIds)

names<-featureNames(diffGenes)
row.names(names)=names$tracking_id
diffGenesNames<-as.matrix(names)
diffGenesNames
diffGenesNames<-diffGenesNames[,-1]
diffGenesNames

diffGenesData<-diffData(diffGenes,x='B73',y='B74')
row.names(diffGenesData)=diffGenesData$gene_id
diffGenesData<-diffGenesData[,-1]
diffGenesData

...

Thank you in advance for the help.

[muthu545]

Hi,

use the following to get the desired output.

diffGenesData<-diffData(genes(cuff),x='B73',y='B74')
sig.diffGenesData<-subset(diffGenesData, significant=='yes')
sig.diffGenesData

Note: The DiffData works on the initial cuffset object 'cuff' and not on a filtered dataset as in your case.

Thanks

[Enrique Zudaire]

Hi muthu545

Thank you, this works great

[RemULg]

Hello to both and others,

I'm a quite new R user and after using the same Tophat/cufflinks/cuffdiff/cummeRbund, I also face a problem with diffData().

I can perfectly use the diffData() function but I would like to invert sample_1 and sample2, so that the log_2_fold_change is calculated with value_2 as the reference value instead of value_1

Any idea of how to proceed?

Thanks in advance for your help!

[muthu545]

Hi,

The best easy solution is to export the results to excel and do an inverse of log2FC column.

Note: Tweaking cuffdiff object is bit tedious.