@Thejas : My protein is not a kinase and hence the process does not involve ATP

Often, such motions are linked to binding events, so the energy involved has to be low enough that they can be powered by these events, that is, lower than the interaction energy between protein and ligand.

@Annemarie Honegger : Dear Ma'am Thank you so much for the response. My protein of interest a transporter of lipids in the plasma. Therefore can I attribute this hinge movement to the protein's intrinsic dynamics?

The larger point here is to what extent the bending and twisting take place in the presence and absence of the lipid. Proteins exist as an ensemble of states, and binding of ligands is generally viewed to shift the populations that are sampled, so energy barriers are indeed affected. The hinge movement is likely intrinsic to the protein, but the better question is how it is impacted by the binding of the lipid and that can only be demonstrated by studying both apo and bound states and their respective conformational ensembles.

@Justin : Since my protein is transporter of lipids in plasma, unless it is attached to a membrane, it shall never exist in apo form. In that case, is it acceptable to remove the lipids from the protein and perform the simulation when I study the protein in plasma (ie water box)? When a conformation is not physiologically possible, are we right in simulating such a conformation?

I don't quite understand. Does the protein exist in the plasma (soluble) or is it bound to the membrane (integral or peripheral)? A distinction needs to be made between the lipids it is transporting and the lipids with which it may make contact. If it is indeed a membrane protein, simulating it in a box of water is not sensible.

The more detail you can provide, the better.

Are you asking why energy is necessary to accomplish these movements? If so, then generally the answer is steric clash from either solvent shells, breaking ion pair interactions or hydrophobic interactions, or steric clash from the two subunits themselves.

But if you're asking where the energy comes from to overcome the barrier, then Annemarie is spot on. It's usually binding energy from protein protein interactions. Or, if the barrier is low enough, it could be simple thermal motion.

Since you've already stated your protein does not use ATP (or other NTP?) then I assume it isn't derived from some reaction mechanism.

@Annemarie Honnegar and @Jason Fowler: Since I am simulating a protein that is a pseudo dimer and carries has ligands- I am obliged to first distinguish which of the two (either interactions between residues at the interface or ligand binding) is causing it. Is there a way to computationally calculate the energy state of the protein in these conformations?

@Jason Fowler : My question was the second part-- where does the energy for this transition come from. Neither ATP nor any other NTPs are involved in this process.

For many such motions, different conformational states are thought to co-exist in a dynamic equilibrium in solution, thermal energy being sufficient to overcome the barriers. A ligand binding just to one state will draw the equilibrium to that conformation, because the interaction lowers the total energy of the complex. Therefore, a conformation that in absence of the ligand is only populated extremely sparsely may become the dominant population in the presence of the ligand

"Since my protein is transporter of lipids in plasma, unless it is attached to a membrane, it shall never exist in apo form."

If this is indeed the case, it may either be due to binding interactions between the transport protein and the membrane, or simple to the energetics of ligand transfer. Of course, the ddG for a lipid to transfer from the protein bound form to water is significantly different from the ddG for the transfer from protein to the lipid bilayer. So the need to be attached to the membrane to release the lipid ligand may not necessarily be due to a conformational change upon binding to the membrane, but to the dfferent ddG of the transfer. On the other hand, to allow such a direct transfer to the membrane, the protein has to associate closely with the membrane (or with membrane embedded proteins). This binding energy to the membrane or receptor may help to stabilize an alternative conformational state that facilitates unbinding of the lipid.

"...is it acceptable to remove the lipids from the protein and perform the simulation when I study the protein in plasma (ie water box)? When a conformation is not physiologically possible, are we right in simulating such a conformation?"

If the conformation is physiologically impossible, it is meaningless.

Any simulation involves simplifications, and you have to analyze your simulation carefully to distinguish between real effects and artifacts introduced by your simplifications. A simulation is only meaningful if you can relate the transitions you observe in the simulation to experimental observations.