How to get information about promoter from bulk-RNAseq?

Question

[camillab.]

Hello!

It my be very very naive question (so apologies if it is) but I was wondering if there is a package/s to extract information about the promoters (e.g., which promoters drives which genes, if multiple genes are regulated by the same promoters etc) from bulk RNA-seq dataset (and/or potentially ssRNA-seq) and not CAGE dataset?

The idea would be to identify promoter/s that activate only specific gene/s starting from published bulk RNA-seq dataset and verify the cell specificity in ssRNA-seq to design cloning vectors to test in cells/tissues (ultimate goal).

Apologies if this is not the right place to ask this question and feel free to cancel the post if it is not!

Thank you in advance
Camilla

[rpolicastro]

Generally speaking, most promoters only initiate transcription for a single gene, with a single gene potentially having multiple promoters driving multiple isoforms. Do you perhaps mean enhancers? Enhancers can drive or repress the expression of multiple genes.

[camillab.]

yes! clearly the terminology on my side needs to improved!

Answer 1

[rfran010]

What type of RNA-seq data do you have?

With polyA, I think it will be hard to incorporate enhancer info without additional datasets.

With ribo-depleted libraries, it may be possible to examine RNA read counts over known/predicted enhancers and compare to nearby genes. However, I do not think you could reliably find specific enhancer/gene interactions, but at the least you find find candidate active enhancers?

[rfran010]

I'll just add what I did with my own dataset of ribo depleted RNA-seq:

I took predicted enhancer-promoter interaction datasets from UCSC and counted RNA reads over the enhancer regions. I also had standard counts over actual genes. Then using the predicted gene partner for each enhancer, I correlated the Enhancer and Gene data (e.g. x-axis was counts over enhancer region, and y-axis was counts over the predicted gene partner for each enhancer region).

I saw moderate (~0.5) correlation. Meaning if an enhancer region had increased RNA, it's supposed target had a good chance of also having increased expression.

Answer 2

[Trivas]

I'd probably select the genes of interest from your bulk RNA-seq results then use the UCSC table browser to get the sequence upstream of the transcription start site (approximately equal to promoter sequence). Note, this is leaving out any effect enhancers or distal promoter elements will have on gene expression.

You can also use a bed file and bedtools getfasta to streamline the process a bit and make it gene identifier agnostic.

[camillab.]

and what about enhancer? is it possible to do it from a bulk-RNAseq?