Hi, is it possible to do transcript variant analysis on salmon output file and which file would be used for variant analysis?
If you mean isoform level transcript DE analysis you can use Sleuth (https://github.com/pachterlab/sleuth), accessible through the wasabi package (https://github.com/COMBINE-lab/wasabi). If you mean splicing analysis, it is better to use isoswitchanalyser in my opinion (https://bioconductor.org/packages/release/bioc/html/IsoformSwitchAnalyzeR.html).
Thanks Chris, I will check them out and yes I meant splice variant analysis. Since I have used SALMON to analyse differential expression and it did not generate a .bam file which I suppose is required for the splicing analysis. I do have count data one of the output. So can I do splice variant analysis with that?
Yeah you don't need the BAM file anymore, you can just use the salmon counts with isoswitchanalyser.
If you want new gene models you can generate them using hisat2 and stringtie or strawberry, this is a lot more work and a different pipeline. People do this if they think their annotation is not that good, i.e. a new species with poorly annotated genes. Maybe there are some diseases and tissues in humans where the annotation isn't that good, I don't know.
Thank you for your time.
Can you say more about your goals here? I'm not sure what you mean by variant analysis.
There is a Bioconductor workflow, rnaseqDTU, for differential transcript usage.
This workflow makes use of the methods DRIMSeq and DEXSeq to perform statistical analysis of DTU:
We have done rna-seq on samples treated with inhibitor of a class of chromatin remodellers. Our chip-seq data showed some of the remodellers bind to 1st intron of genes therefore we want to check whether inhibitor is mediating alternative splicing by affecting chromatin remodellers. Hope I am a bit clearer.
Ok, so then the above workflow should work for you to detect alternative splicing, and we go directly from Salmon output.
If you have future questions, feel free to tag with "rnasetDTU" so I will get a notification.
Thank you and sure.
Login before adding your answer.
Use of this site constitutes acceptance of our User Agreement and Privacy