Entering edit mode
4.9 years ago
sabaghianamir70
▴
70
Hello
I want to detect pseudogenes from my result list. I got my result from Deseq2, but i can get it from egdeR or limma. is there any tool or any way to extract them all easily ?
I'm not sure where you're going here?
You want to get pseudo genes out of your expression analysis? Or do you have a list of pseudogenes you want to filter out from the results obtained?
In any case being pseudogene is not something you can detect from expression analysis, it's a "structural feature" of genes, not linked to their potential expression.
What do you mean i cant detect in expression analysis.. I have been detecting them like: Pseudogene(gene) GBP1P1(GBP1) and PER4(PER3) in my results, but i want to to extract them all from my analysis.
I think there is a misunderstanding here. lieven.sterck thinks that you want to detect pseudogenes (without knowing if a gene is classified as such) simply based on the DEG results (so a p-value, or a certain fold change). I think you already know the classification of your genes and simply want to pull out all genes from your DEG that are classified as such. Is that correct, if not please elaborate by giving a representative output example.
thx ATpoint .
Indeed, I (falsely?) was under the impression you want to de-novo detect/identify pseudogenes from your DEG results. If that is not the case (== you want to filter) , then something along the lines of what ATpoint indicated here C: how can i detect pseudogenes from my Deseq2 result ? is likely the best way forward.
I assume you have done DEG and now want to extract those genes classified as
pseudogenein existing annotations, right? In that case I would get an annotation file (GFF or GTF) for your species, then extract the genes annotated as pseudogene. Then simply overlap those gene names with your DEG list and select those from your list that match. Not aware of any tool to do that but could be done with 'awk' or a few lines inR. If you need more specific help please provide code you've tried and we can try to debug if necessary.Thank you, it can help me.
For to be sure about this
this is my Deseq2 result. those Gene ID are transcripts which some of them are genes and some of them are pseudogenes. Now i want extract genes which annotated as pseudogenes, So based on this, the solution ATpoint mentioned is correct ?
Yes, you could now filter for those annotated as pseudogenes. Still, are you sure you want to do the DEG analysis like this? What I mean is that DESeq2 is supposed for gene level rather than transcript level analysis. Still, you perform differential analysis apparently on transcript counts. You typically summarize your transcript counts to the gene level prior to DEG. Check for example the
tximportpackage which does exactly this. I suggest you read (and potentially exactly follow) the DESeq2 manual and the RNA-seq workflow from its developers (https://www.bioconductor.org/packages/devel/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html).Adding on this, there is also the
tximetapackage (https://bioconductor.org/packages/release/bioc/vignettes/tximeta/inst/doc/tximeta.html) which essentially does whattximportdoes (so summarizing transcripts to the gene level) but it also pulls metadata for common organisms, such as human. This would (I guess, never used it) provide you right away with the information towards a gene being classified as pseudogene. Check out its manual, might be easier to use this than custom filtering.Its because i put my own reference genome in GALAXY. If i using the default reference genome the transcript ids shows as Ensembl IDs. The GALAXY.eu default is based on h19, but i put the h38 version of reference genome, is this wrong ?