We are going from protein sequence to protein structure. Does MD simulation refine the protein structure? if we have to perform MD simulation then do we have to perform loop modelling and protein refinement steps as well?
That depends on how ”good” your starting structure is. If you have all atoms present and there aren’t many clashes then you probably don’t need to refine too much. Some force fields/simulation software handle clashes better than others. You can always run preliminary simulations to see if you get reasonable results without refinement. You need a property you can calculate computationally that’s also known experimentally to compare to and validate. If your calculation is within reasonable range of the experimental value, then your model is probably “good enough.” Otherwise, you would want to refine. A good way to refine a model is to use ISOLDE, a ChimeraX plugin, that fits a
model to an x-ray or cryo-EM map. But there has to be an experimental structure for that. The software and tutorials are online for free.
Crystal Structure: While doing MD simulation, it is necessary to prepare your raw protein structure which is downloaded from the PDB database. In case of missing atoms, side chains, and/ missing loops you must have to perform protein refinement and loop modeling.
Modelled Structure: In this case, you must have to refine protein structure and loop modelling to get a stable and energetically minimized model and then you can perform the MD simulation.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators