This is a test version of Biostars. For the public version, visit https://www.biostars.org.
differential expression with cuffdiff

Hi all. I have 24 samples RNA-seq, paired-end data including 2 genotypes, 3 Zone from root, 2 conditions and each condition including two replicate. How can I do cuffdiff for this expriments. thanks alot

rna-seq
• 3,146 views • link
updated by Kristoffer Vitting-Seerup Kristoffer Vitting-Seerup • written by yaghoub.amraei • yaghoub.amraei

2 answers

Cuffdiff is no longer recommended, update your analysis to HISAT / StringTie / Ballgown

• 0 views • link
written by JC JC

Thanks, Can't use Caddiff with this number of samples?

• 0 views • link
written by yaghoub.amraei • yaghoub.amraei

Yes, you can, but again, that is deprecated

• 0 views • link
written by JC JC

The authors of those tools say so themselves (see their lab websites; their twitters; their previous posts on biostars; etc.).

• 0 views • link
written by dsull dsull

I have not been able to find it. Could you point to something more recent than this (which states cufflinks it is not deprecated).

• 0 views • link
written by Kristoffer Vitting-Seerup Kristoffer Vitting-Seerup

Sure.

Most recent: This mentions it as "retired software" and "no longer under active development and are not being supported". It does not list it as "now maintained/developed elsewhere".

Older posts: See this tweet (referring to it as obsolete). And see this.

• 0 views • link
written by dsull dsull

I would also not use this pipeline. Stringtie is overly complicated unless you want to assemble the transcriptome. For a standard DE analysis (i a proper reference transcriptome exists) there is (imho) no need to bother with it. ALso leightweight pseudo- or selective aligners such as kallisto and salmon typically outperform traditional alignment for RNA-seq quantification (check any of the recent benchmark papers). This ballgown tool does not really provide features which other, better documented and actively maintained tools such as DESeq2 or edgeR do not offer.

I would go with quantification by salmon, importing data into R with tximport, then DESeq2, as described e.g. here: https://www.bioconductor.org/packages/devel/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html

• 0 views • link
written by ATpoint ATpoint

Unfortunately you cannot analyse your experimental setup with Cuffdiff as it cannot handle overlapping conditions/co-factors. For this you need to use cuffquant and cuffnorm (see overview here) to get a count matrix which you can then analyze with DESeq2, edgeR or similar.

• 0 views • link
written by Kristoffer Vitting-Seerup Kristoffer Vitting-Seerup

Can you explain more about this or give an example?

• 0 views • link
written by yaghoub.amraei • yaghoub.amraei

Log in to answer this question.

More posts like this