CHARMM Development Project Forums User Discussion & Questions Molecular Dynamics Protein/metal pulling simulation

Howdy all,

I've run a series of simulations where, having found an opening, a metal and counter-ion is pulled from outside the Transferrin C-term binding pocket into where the metal binds to some residues, traversing about ~16A through a sort-of 'channel'. I've run these simulations for three metals (Fe, Eu and Cm) and varied the magnitude of the pulling force, with and without the counter-ion, etc.

There appears to be differences in the ease with which different metals traverse the channel. Eu and Cm are much bigger so, well, duh. Eu is a harder acid than Cm so you'd imagine there would be difference in binding to polar residues at the end of the trip but, in this case, I'm gathering that ionic radius is the bigger factor at play here.

What I'd like is to tie what I'm seeing to something measurable experimentally, ideally, but a meaningful quantity to make the metals comparable is the main goal. The channel is largely made up of apolar residues and you'd imagine the system would pass through a number of very similar microstates so I don't think a free energy calculation gets us there. I know CHARMM allows one to calculate a diffusion coefficient, which from what I can see is done via post-processing of the trajectory (true?). The latter would seem to most likely way forward.

For such a system, are other quantities CHARMM is able to calculate that might add something else? Doesn't have to be from post-processing, definitely not averse to setting up more simulations.

Aren't you putting the cart before the horse? First decide what quantities you want to calculate, then we can try to help you figure out how to do this in CHARMM.

What I'd like is something that tells me something about the difficulty, or otherwise, of these metals reaching the binding pocket. Iron does it naturally but I suspect Curium gets stuck part way. So if I apply a strong enough pulling potential to force Cm to reach the pocket, I'd like some indication whether the barrier is likely to be prohibitively high in the natural system for this event to occur.

So aside from the diffusion coefficient and deltaG, is there another method within CHARMM that could provide that measure of difficulty of traversing the 'channel'?