I have a list of transcription factors from a high-throughput experiment. I would like to know about the probable target genes of these transcription factors. For example using OregAnno, I found that target gene of KLF6 is LTC4S. OregAnno seems to be not up-to-date. Is there any other bioinformatics resource where I can find the human TF->Target information ?
Tough question. Most people want to go from gene to TF regulating that gene instead of TF to all genes under its (partial) control. What also makes this tough are: 1) tissue-specificity, 2) predictions of one or few binding sites when in vitro/in vivo data would suggest that the particular TF needs to bind multiple times, typically in cooperative fashion, in order to promote transcription. Other complications are different mRNA isoforms under control of different, perhaps overlapping TFs and genome variation altering regulation by a particular TF.
I know that none of this points you to a database of TF-gene interactions. I know of none that would address the above points - and those are relevant to the way I look at this question.
I don't know but I doubt they consider these items. The real deep detailed stuff I have seen puts focus on just one TF. But one of ~2000 certainly won't cut it for your work.
Larry: Thanks for your suggestions on why this is difficult. I am wondering whether or OregAnno or other database that reports target information consider these important aspects.
I recently used "Investigate Gene Sets" options in MSigDB to find the transcription factor target information of my dataset. Adding this here as an answer for future googlers.
The data is curated from TRANSFAC, you can do a 'Compute Overlap' of your genes with the list of genes in the dataset. Output includes a p-value to indicate statistical significance and a Gene/geneset overlap matrix to visualize the gene categorized into different transcription factor targets.
I agree with Larray on that people usually try to search for TFs acting on the interested genes. I have two suggesions to solve this problem. First, we can do it as usual, or find the TFs for all the genes in the genome (if available) of your research species and then select those genes posessing your TFs.This poissibly will take much time for computation. Furthermore, this probably has been done for model organisms and relavant databases may be available.Replicate the procedure and apply it to your own species.MAPPER can help do this. Second, try to find some ChIP data about your TF.If you are lucky and such kind of data is available,it will be a great help.The two suggestions may turn out to be useless since they are based on some preconditions that may be unavailable.
There is also the Transfac database which is a commercial database of TF and interactions with subscription (now, I thought they had had a free portion before?).
As a wet-lab approach the exact solution to your question would be ChIP-sequencing (or ChIP-chips).
If you are happy to go with predictions have a look at the MEME suite http://meme.sdsc.edu/
Just be aware that you are getting plenty of false positives. To get a more accurate picture for a particular scenario have a go with ChipSeq.
I am working in the same.I have a list of Transcription factors, their target genes and miRNAs all involved in skin cancer. Actually I want to make a directed regulatory network. But the problem is how to make it direct because all the databases mentioned above not provide information about what kind of regulation exist? weather it is activation or inhibition or coactivation like that. Manual annotation seems very difficult for me as I have huge data. Please suggest any database, plugins or method so that I can get such type of regulatory information.
i think by reverse engineering tools such as ARACNe you can define Targets Of A Transcription Factor