Hello, my name is Hamid, I have constructed a condition independent gene co-expression network. There are some transcription factors in modules of this network. The function of these transcription factors are in accordance with the module function, but their binding sites were not over represented in that module. I would be very happy if somebody help me to find an explanation for this problem?
Vertebrates known transcription factor binding motifs were downloaded from the JASPAR database. Overrepresentation of known transcription factor binding motifs in the CIS regions of the genes within each module was performed using Clover.
Thanks in Advance
In what genomic region did you search? Searching TF binding sites using sequence motifs is nice in theory but usually doesn't work in practice since the search space is large and they're found everywhere.
Dear Asaf
Thanks to your response, I constructed Soy bean scale free gene interaction network using integration of gene expression datasets derived from diverse range of microarray experiments, including different tissues and experimental conditions.
As far as I know, these genes are co-expressed under different conditions and different tissues. So if a transcription factor included in the network module, its target should be included too. Do you agree?
Not necessarily, it might have a secondary effect and not drive the expression of the module. I'd try motif searching in promoters of the genes in the module as well (using MEME for instance) compared to other modules (discriminating search), maybe you'll get a common motif and find the main TF (though I doubt it).
Can you please explain more about
Dear Asaf thanks to your comments. they were very useful for me. Regards, Hamid
It means that the regulator might not influence the genes in the module but genes in another module.
Dear Asaf
I find the answer of my question. many thanks to your help.