Hello Amber users,
I have a question related to simulating membrane proteins with Amber. I read the Amber lipid tutorial, but it wasn't very informative because it mainly focuses on simulating the membrane itself and the membrane protein example is pretty anecdotal and refers to the membrane only simulations in the beginning. However, having the protein buried in the membrane necessitates other kinds of controls and considerations.
Papers related to the force field itself too focus on the models, simulation of the membranes, and validating the methodology of parameter generation. Although there are papers in the literature related to protein/membrane simulations, they usually have one of these problems that make them hard to compare:
1) They are not run on Amber.
2) They just keep saying "as explained in our former paper" for a series of papers.
3) They account for the overall procedure, but selections of procedures vary from paper to paper and are often vague (which is partly understandable because they don't write for educational purposes).
Overall, I'd like to have opinions of people on practicalities of such simulations. I have a membrane protein which has an extracellular and a transmembrane domain. I put in a POPC membrane large enough to cover the system using CHARMM-GUI and followed the steps until I got my structure that contains protein + membrane + water + ions. Everything looks fine so far when I inspect the system. My questions related to the rest of my case are the following:
1) Once the structure is obtained from CHARMM-GUI (as suggested in the lipid tutorial) in its solvated and neutralized form, next thing is changing structural features are changed to fit Amber standards (HIS names, some atom names, etc).
At this point, is there a need for adding/removing waters or ions, or can we directly proceed to loading the structure and generating the parameters using leap?
2) Some lipid molecules seem to swing from the sides of the membrane. Does this necessarily indicate a problem, or does the periodic box handle this? VMD obtained box information seems to show a slice of the water/membrane separated from the rest following minimization. Should the box encompass the exact size of the system, or is it OK to have a larger box to be prudent?
3) I typically run a heating step with restraints on the protein backbone when it's just a protein in water. But when I take my structure for this system from the minimization steps and restrain the protein only, lipid structure just falls apart during the heating steps. It's clear that it should somehow be restrained, but what's a good way of ensuring both the protein and lipid are heated safely and properly?
Sorry for the long message, but I thought people could benefit from the details in the future.
Best,
Alican
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Apr 20 2017 - 09:00:02 PDT
