Transcription factor interactions via multiple motif discovery

[says_anova]

Hi All,

I have a high confidence ChipSeq data set for TF1 and have run MEME to discover the most likely canonical motif (motif1), which occurs in about 60% of my peaks. So far so good.

Now I would like to generate hypotheses about other transcription factors that may be physically interacting with TF1 to regulate some subset of the genes TF1 regulates. Naively, I would guess that I need to
a) select peaks that contain motif1
b) mask motif1
c) re-run MEME to discover motif(s) of some other TFs that are in close proximity to motif1.

In parallel, I can also
a)select peaks without motif1
b) run MEME to discover which TFs have motifs in those peaks and hypothesize that TF1 physically interacts with those motifs without binding to the DNA itself.

Obvious pitfalls?

Also, I've been googling around to find a tool that might automatically look for paired motifs but haven't found anything yet. Suggestions?

More generally, how would you go about discovering TF cooperation from ChipSeq data for a single TF?

Thanks in advance!

[says_anova]

No one ?

[Ayyappa_kumar]

From your query I understand that you are interested in finding the secondary motifs present in your data. If so you could use SPAMO for such analysis.
http://meme.nbcr.net/meme/cgi-bin/spamo.cgi