I'll preface this by saying our NGS Core facility handles most of our analysis. So perhaps in my limited understanding of the analysis workflow, I'm not correctly articulating what I'm looking to do. Our core has been helpful but also has limited ChIP-seq experience and they aren't sure what I want to do is possible.
We are have run several ChIP-seq experiments seeking to identify where the Aryl Hydrocarbon Receptor binds in the mouse genome is response to various agonists. The current analysis workflow has produced a list of the most common motifs identified, the genomic region associated with peaks (promoter, intergenic, intragenic), and the prevalence of each motif in the dataset. This is useful, however, I'm interested in specifically which genes are located near these motifs. A specific motif to be precise. Our lab has previously reported a non-canonical binding site which consists of the tetranucleotide repeat GGGA. I want to know the location of peaks in the ChIP-seq dataset that correspond to this motif. Preferably the location of peaks corresponding to this motif within gene promoters (within 5kb of the TSS I suppose). The goal is to identify putative, non-canonical target genes. Is this possible?
Basically, we have a list of peaks, distribution of regions associated with peaks, and a list of specific motifs. How do I cross-reference them to find a specific motif that corresponds to specific peaks at specific regions (promoters)?
I think you may be able to come up with a consensus binding site sequence using ChIP-seq localization data for your receptor, and then to see how the presence/absence of such sequence motif(s) correlates with activation of transcription in response to various agonists (RNA-seq data).
Eric,
This should be a relatively simple task for somebody who is comfortable processing NGS data. One should easily be able to identify peaks corresponding to your unique non-canonical sequence (and excluding any overlapping consensus sequences) and determine their position in the genome. This can then be associated with nearby annotated genes without much trouble at all. I would ask your core if they can do this for you or find somebody with bioinformatics experience and they should be able to help quickly.
Jeffrey Nicholas Mcknight
Thanks, I'm glad to know what I want to do is feasible.
I have asked our core if they are capable of this and seemed unsure. They have much more experience with RNAseq analysis than ChIPseq as not many labs at our institution seem to run ChIPseq. Are you aware of any specific software, tools or packages I could recommend our core look in to?
Eric - Unless I'm misunderstanding, you already have the genomic locations of where your protein is binding. If you already know the locations of your noncanonical motif of interest, you should be able to simply extract nearby genes from a gtf file (I would personally write something in Python to do this, but somebody better than me might have other resources). You may also want to consider other nearby relevant features (eg enhancers) depending on what you might be expecting.
- Badges
- Science method