Tool for transcription factor driven prediction of binding sites

Question: Tool for transcription factor driven prediction of binding sites

5 months ago by

Hi everybody!

I have a fungal transcription factor, and I suspect it is a negative regulator of a certain set of genes: do you know any tool that might help me look for binding sites in a transcription factor driven way?

All the tools I am finding work on the promoter sequence of the target genes without taking into account the actual sequence of the transcription factor...

May those who answer me be rewarded with pizza and cake.

binding sites promoter transcription factor • 232 views

ADD COMMENT • link •

modified 5 months ago by Alex Reynolds ♦ 28k • written 5 months ago by edoardo.piombo • 20

5 months ago by

Alex Reynolds ♦ 28k

Seattle, WA USA

Alex Reynolds ♦ 28k wrote:

You could grab the sequences of your target gene promoters and run them through MEME to find putative binding sites (basically, use something like e.g. bedops --range -1000:0 and bedops --range 0:1000 to get BED files of promoter regions from stranded genes, convert the BED to FASTA via samtools faidx queries and a helper script, and run the FASTA through MEME), and then use TOMTOM to compare the search result against your motif model(s) of interest.

ADD COMMENT • link modified 5 months ago • written 5 months ago by Alex Reynolds ♦ 28k

Hi! First of all thank you for the suggestions, I'm doing all of that, but I have a question: how do I determine which is my motif of interest, starting from the sequence of my transcription factor? Sure, I can use MEME on the promoter sequences and then use TOMTOM to compare my motives with various databases, but is there not a way to understand what motif I am looking, for starting from the transcription factor sequence?

ADD REPLY • link written 5 months ago by edoardo.piombo • 20

Maybe I misunderstand, but having one sequence is not enough. You need multiple sequences to build a motif model.

If you have multiple DNA sequences that represent your TF binding sites of interest, you could run MEME on those, to get back a model.

You could take that model and the MEME models that come back from searches on your promoters, and then do a similarity search on them via TOMTOM, to see if you get any significant matches between your putative TF and what's in your promoters.

ADD REPLY • link written 5 months ago by Alex Reynolds ♦ 28k

Ok, so maybe I did not express myself well: I have no idea about which are the binding sites of my transcription factor. I know it is a putative transcription factor by blast alignments and interproscan analysis, and I suspect it is involved in the regulation of a specific set of genes due to its position near these genes in the genome. What I was wondering about is this: is it possible to predict the binding sites from the coding sequence of the transcription factor itself?

ADD REPLY • link written 5 months ago by edoardo.piombo • 20

I wish I could help. I don't know much about using protein structure or sequence to predict TFBS. Some papers that might link to other papers of interest: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4908348/ https://www.ncbi.nlm.nih.gov/pubmed/28715997

ADD REPLY • link written 5 months ago by Alex Reynolds ♦ 28k

Thank you! I will check them out!

ADD REPLY • link written 5 months ago by edoardo.piombo • 20

Similar posts • Search »