2
$\begingroup$

I have the counts matrix of an RNA sequencing dataset of fibroblasts and I wish to identify isoforms of a particular gene of interest in it. Can anyone please hint me on a bioinformatics method to achieve this?

In literature, 9 different isoforms of my gene of interest have been reported. My colleague who does the wet lab experiments has identified 3 of those isoforms through western blotting in the fibroblasts that later got sequenced. There might be more isoforms in these fibroblasts.

I will be grateful for any computational hint to identify gene isoforms from my dataset, irrespective if they are differentially expressed or not.

Thank you in advance for your kind help.

$\endgroup$
  • 3
    $\begingroup$ What tool did you use for quantification? Most tools produce transcript level counts, so this should be a pretty straightforward task once you get a list of the identifiers of the transcripts you're interested in. $\endgroup$ – Ram RS Mar 24 at 16:32
  • 1
    $\begingroup$ If you have only counts on a gene level, you might need to go back to the bam file to get the isoform quantification.. Something quick might be kallisto? $\endgroup$ – StupidWolf Mar 24 at 18:36
2
$\begingroup$

You can back to your bam file and use cuffdiff tools it is very able to do your job follow this to liks http://cole-trapnell-lab.github.io/cufflinks/manual/

https://github.com/Jeanielmj/bioinformatics-workshop/wiki/Differential-Analysis-with-Cuffdiff

My best

|improve this answer|||||
$\endgroup$
  • $\begingroup$ As a remark, since it is 2020 you better you a recent and non-deprecated tool such as salmon or kallisto to get transcript level abundance estimates and then do differential transcript expression analysis with tools such as DEXSeq, swift or sleuth (the latter is downstream of kallisto from the same group). $\endgroup$ – ATpoint Mar 25 at 20:03
1
$\begingroup$

If literally all you have is the count info by gene, you can't magic up the isoform breakdown.

|improve this answer|||||
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.