Cuffdiff only giving gene names to certain sequences

Question: Cuffdiff only giving gene names to certain sequences

8 weeks ago by

lcrisman • 0 wrote:

Hi everyone,

I have been trying to figure this out for the better part of a week so hoping someone has had the same issue. I am using the cufflinks -> cuffmerge -> cuffdiff pipeline for RNAseq data. For cuffmerge I used both the mm10.gtf reference file as well as a genome .fa file. The output from cuffdiff is only giving gene names to comparisons that give a NoTest status. Anything that was tested wether significant or not has no gene name given to it, obviously I am most interested in significantly different gene expression so the information I want most is not being given. I do not know what would make the program give notest transcripts the matching gene name and actually tested transcripts no name. Please let me know if anyone has a solution to this or even just something I could try as right now I am completely out of ideas.

Thanks in advance!

rna-seq cufflinks • 76 views

ADD COMMENT • link •

modified 8 weeks ago by Jennifer Hillman Jackson ♦ 25k • written 8 weeks ago by lcrisman • 0

8 weeks ago by

Jennifer Hillman Jackson ♦ 25k

United States

Jennifer Hillman Jackson ♦ 25k wrote:

Hi,

"NOTEST" means that there wasn't enough data clustered for the tool to make a reliable match to a known transcript/gene (or to even include it in a novel transcript/gene, unless you used the tools in a way or set the options in a way that filters discovery out). http://cole-trapnell-lab.github.io/cufflinks/cuffdiff/#differential-expression-tests

You could isolate these regions and compare them directly to the reference annotation (overlapping coordinates) and review for yourself or pull the data up in a genome browser and visualize/review the regions. Please be aware that something else could be contributing to the problem. The setting used with this and the prior tools may be too strict. Or, maybe too much QA was done that removed reads that you could have kept? Or there was a mapping problem (check the fraction of reads that mapped concordantly - the goal is to maximise this). The possibility of a chromosome mismatch problem between the GTF and the custom genome fasta could also be a factor. FAQs: https://galaxyproject.org/support/#troubleshooting

That said, Tophat and all of the Cuff* tools are considered older and deprecated now. There are known bugs with a few of these tools (that wouldn't be expected to impact your results if you got this far already -- but that can't be ruled out). It is strongly recommended to move to the current methods and tools. The Galaxy Tutorials cover example RNAseq workflows/tools: https://galaxyproject.org/learn/

Start here: RNA-seq: Discovering and quantifying new transcripts - an in-depth transcriptome analysis example. https://galaxyproject.org/tutorials/nt_rnaseq

Thanks! Jen, Galaxy team

ADD COMMENT • link written 8 weeks ago by Jennifer Hillman Jackson ♦ 25k

Similar posts • Search »