Question: Cuffdiff-differential gene expression-NOTEST
gravatar for raj_262920
3.8 years ago by
raj_26292020 wrote:


I am just starting up and getting to know and analyze my data from the RNA-seq. I have 2 biological replicates for 2 different conditions, Pair-end sequenced. The workflow is as follows: I ran FastQgroomer on all my reads, then I used these reads to align to a reference genome using Tophat2. Tophat 2 analysis showed about 80 % alignment with the reference. Then I ran Cufflinks on the accepted hits from Tophat2 with the reference genes and genome. I verfied the cufflinks output files and I can see FPKM values for the two different conditions and replicates. After this i used cuffmerge to merge the transcripts from all the cufflinks outputs. so i merged the four assembled transcripts files from cufflinks( 2 conditions and 1 replicate for each conditions). Then i ran cuffdiff to find differenctial expresssions and so on, using the merged transcript file from cuffmerge and i also used the reference genome. In the cuffdiff output..differential gene/transcript expression tracking files..I see the FPKM is totally zero and i see 'NOTEST' for all the samples.

What does that mean ? I dont think that means that there are no significanlty expressed genes between the samples ! . When i compare the FPKM values from cufflinks, there is big differene in FPKM values for the 2 conditions !

What can be done to solve this ? Please suggest !


Thank you and Best Regards


rna-seq galaxy • 3.2k views
ADD COMMENTlink modified 3.8 years ago by Jennifer Hillman Jackson25k • written 3.8 years ago by raj_26292020

I am having the same problem and I am trying to use the .gtf annotation file with the merged.gtf file from Cuffmerge on the Cufflinks. Does someone knows how to use both together in a workflow???

ADD REPLYlink written 2.8 years ago by juafonso_bio0
gravatar for Jennifer Hillman Jackson
3.8 years ago by
United States
Jennifer Hillman Jackson25k wrote:


This indicates that there is not enough data to detect significant differential expression between the samples. Using a reference annotation dataset (GTF, GFF3) with Cuffmerge could help group the data together. Use the iGenomes version, if one is available for your genome. More here:

Thanks, Jen, Galaxy team

ADD COMMENTlink written 3.8 years ago by Jennifer Hillman Jackson25k

Hi Jennifer

Thanks !  So you meant that i should use in cuffmerge , the first transcript file (GTF file produced by cufflinks), could be the gtf annotation file from my reference ? and my other cufflinks generated transcripts as additional GTF files ?

If so, i should say 'no' to the option 'use reference annotations' ? or did you actually meant to use a reference annotation file here ? I actually used one before !


Thanks so much !

ADD REPLYlink modified 3.8 years ago • written 3.8 years ago by raj_26292020

On the tool form, enter the Cufflinks output files for the option "GTF file produced by Cufflinks:", then also use the option "Use Reference Annotation:" as "yes" and include it there when the form resets. Hopefully that helps, Jen, Galaxy team

ADD REPLYlink written 3.8 years ago by Jennifer Hillman Jackson25k

Oh, I didn't see that you had already used an annotation dataset. The data could just be too sparse, or that could potentially be contributing to the problem in Cuffdiff. Double check that the chromosome identifiers in the reference annotation file are an exact match for those in the reference genome used. If they do not match (are not based on the same exact reference genome), many issues can arise.

ADD REPLYlink modified 3.8 years ago • written 3.8 years ago by Jennifer Hillman Jackson25k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 175 users visited in the last hour