Simulating 2 proteins with go-like model

[Pbagchi05]

I have a protein complex with a Cryo-EM structure, and I am using that as the starting structure for my simulation.
The complex consists of two proteins. I want to know how I should martinize it—should I generate two separate coarse-grained structures and use them independently, or should I coarse-grain the entire complex together?

My main concern is about the CA virtual sites. If I coarse-grain the entire complex together, Go-contacts will be generated between the two proteins. If I martinize them separately, then there will be no contacts between them.

Could you please tell me which approach is correct?

[csbrasnett]

Really, the answer to this is pretty system/simulation goal specific.

If you care about the proteins being in a complex with a particular interface, then it's probably best to martinize the whole complex together in one go. That way, you'll (probably) guarantee that the complex won't fall apart during simulation.

If on the other hand, you want to understand how the complex might dynamically form/interact with its environment/etc with a simulation, then you're right that it'd be better to martinize them separately, and afterwards reconstruct the initial starting structure. In this case, it might also be easier to use an elastic network and take advantage of the -eunit option, specifying that the elastic network should only be formed within a chain.

[Pbagchi05]

Yes, I want to see the dynamics, and also does it bind through the course of simulation, so I think martinizing them separately is what I should do.

Thanks!