Himanshu Grover wrote: > I am trying to model a protein complex, with 2 chains. I wish to know > the interaction energy between the two chains at the end of > model-generation. Is it possible to do this, taking into account both, > van-der-waal's interactions as well as coulombic? Is the energy value, > that appears at the top of each generated model pdb file, indicative > of the same (While optimizing the structure, I have Dynamic Lennard > and Dynamic Coulomb switches on, and others off.) > I am not looking for the internal energy of the whole protein complex, but > just the interaction energy of the 2 comprising chains.

You'd do this in Modeller in a similar way to in any other molecular mechanics package; calculate the energy of chain 'A' on its own, the energy of chain 'B' on its own, and the energy of the complex, then calculate the interaction energy as E(A) + E(B) - E(AB). Whether you decide to optimize any of these structures (e.g. to allow for conformational change on binding) is up to you.

Modeller by default gives you the energy of the whole system, but you can calculate the energy of just one part by using PICK_ATOMS first to select a subregion.

Ben Webb, Modeller Caretaker