Start with asking yourself how to fold a regular protein in PyRosetta/Rosetta. There are very well written tutorials and sample scripts to help get you started. To fold a membrane protein specifically, you could taylor PyRosetta to your needs. Aka, changing the scoring function accordingly, folding certain parts first, etc. Are you trying to fold the entire protein, or just the extra/intra cellular protein? Is there any homology to any known structures for any of the domains you are interested in? With PyRosetta, there is no one way to do anything. Even simple folding. I would honestly start with Rosetta itself, and then once you know more of what you are looking to manipulate, move onto PyRosetta. Hope that helps.
Thanks for your reply.
I have been folding membrane proteins and soluble proteins in Rosetta 3.1 for a while. But I want to move on to incorporate some experimental data during folding. According to my understanding, simply changing the score function wouldn't help. There is a multilayer membrane environment designed to do this job in regular Rosetta release 3.X. What I really need is to have some flexibility in pyRosetta, that would allow me to vary the membrane environment and refine structure with some data. Homology may help. But it will be great, if I can do this in PyRosetta with more flexibility.
"I would honestly start with Rosetta itself,"
To further Jared's point, there is a membrane abinitio mode released with 3.2.