We are developing tools to analyze regulatory network topology in terms of self-consistency and predictive ability. 

Our major tool is the use of hormone-regulated forms of various transcription and regulatory factors, protein synthesis inhibitors, and quantitative RT-PCR (qPCR) based, time-course studies. 

For simplicity's sake, we define the direct targets of a particular protein as those whose regulation (at the RNA level) does not required translation.

Another feature of the use of hormone-regulated forms of proteins is that while stable in the absence of hormone, they tend to be rapidly degraded upon hormone addition [1,2].

This makes the response of an animal cap to activating hormone very much of a discrete pulse, and enabling us to separate the initial perturbation from down-stream effects. 

Dose-response experiment can then be used to define regulatory thresholds involved in different responses.  

We are currently collecting the response from previous published studies, and including them into an interaction data base; we are also incorporating our own time-course studies.