Are There Any Manual Curated Human Tf-Tf Interaction Database?

Question

[michaelchen333333]

I have an ESC-related TF dataset (about 13 TFs), I would like to know if each of the ESC-related TF interacts with any other "TF", so I am looking for a database that has TF-TF interaction data. I search the NCBI and EBI but they don't have such TF-TF interation database yet.

I did find some TF database, such as TRANSFEC (not free...), TFCat (more like a TF database, no TF-TF interaction data).

Any suggestion well be nice.

Answer 1

[Charles Warden]

"TF-TF interaction" seems to imply that you are looking for a protein-protein interaction. It this true? If so,TRANSAC won't be appropriate (it provides position weight matrices for the DNA binding sites).

I think one of the most comprehensive (but also not free) of manually curated interactions in in IPA. You can build networks based only upon protein-protein interactions. They also have TF-target interactions (if that is actually what you are looking for).

There are some free protein-protein interactions databases, and you could see if there are text-mining tools that partially recover the information in IPA (which might be practical, given that you only have 13 TFs):

http://cdwscience.blogspot.com/2013/03/bioinformatics-101-protein-analysis.html

http://cdwscience.blogspot.com/2013/03/bioinformatics-101-literature-text.html

[Charles Warden]

I also noticed this was published today (if you are interested in DNA-binding, not protein binding). However, looks like it has only been tested in Drosophila and it isn't manually curated:

http://nar.oxfordjournals.org/content/42/5/2833.full

[michaelchen333333]

Thanks a lot, I am actually looking for PPI between TFs like you said. I thought there might be some database with such things like they focus on collecting TF interacted with TF data, but It seems like I have to look at PPI databases to find out such things, is that correct?

[Charles Warden]

Sounds like this is the case; luckily, I think you just want a subset of the results (which you can parse).

In the case of IPA, you can create a network among your 13 TFs that shows all known protein-protein interactions. Not sure how easy this is to do with the free software, but the underlying information should be there and I think working with 13 resulting lists should still be manageable.

[michaelchen333333]

Ok. I will try using IPA for a bit, thanks a lot.