This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Cummerbund Gene Tracking

Hello,

I am having an issue with gene tracking in cummeRbund. Here is the cuff output:

> cuff
CuffSet instance with:
     4 samples
     23285 genes
     30073 isoforms
     25872 TSS
     24748 CDS
     139440 promoters
     155232 splicing
     122340 relCDS

But when I try to run gene tracking, the following happens:

> myGeneId <- "ENSMUSG00000045440"
> myGene <- getGene(cuff,myGeneId)
> myGene
CuffGene instance for gene ENSMUSG00000045440 
Short name:     ENSMUSG00000045440 
Slots:
     annotation
     features
     fpkm
     repFpkm
     diff
     count
     isoforms     CuffFeature instance of size 1 
     TSS         CuffFeature instance of size 0 
     CDS         CuffFeature instance of size 0 
> genetrack <- makeGeneRegionTrack(myGene)
Error in `[.data.frame`(features(object), , featCols) : 
  undefined columns selected

Therefore, I'm unable to visualize transcript-level features of my gene of interest. I am using an Ensembl gtf and have set options(ucscChromosomeNames=FALSE), but no luck. I have also tried using the cummeRbund cufflinks data from the sample directory a<-readCufflinks(system.file("extdata", package="cummeRbund")) but still no luck. I have no issue analyzing 'myGene' or any other cummeRbund feature, so I assume I've accessed the data correctly. Any ideas on what the problem might be?

Thanks.

cummerbund r rna-seq

If you post your entire R script as well as explain what you expected to happen, that would be helpful.

have you tried using the recommended genes.gtf file downloable from the Tuxedo package website?

1 answer

You might take a look at this post on SEQanswers. You still did not post your entire R script, as we do not know what your cuff data frame looks like, and I suspect that your issue arises because your are not reading into R the proper GTF output from cufflinks.

@Matt I've now included the cuff output in the post - I think this is what you were looking for.

Yes, for me this was the solution: Make sure to pass "merged.gtf" to cummeRbund.readCufflinks(), if that GTF-file was also used in the previous analysis with CuffDiff.

Log in to answer this question.