Option to a partial modeling
|Opened on:||Saturday, June 26, 2010 - 22:10|
|Last modified:||Saturday, June 26, 2010 - 22:10|
If you plan to expand the set of tools that I have a suggestion
Would be useful to run a partial simulation of not all protein, but only a given part of it.
Ie perform calculations for only selected segments. This is not the same as that local_wiggle or freeze part of the protein and do something with the rest. Now even if you select a certain part, and set everything else to freeze when you start the simulation, all the rest too involved: although it does not move, but the impact on the variable piece. I'm talking about modeling the selected part, as if about nothing else but her - as if the selected piece is the whole protein.
This must be such as to correct some incorrectly folded structures, such as a long helix or sheet and the "break" in the middle. Attempts to fix it often hinders, what he had already tightly interacts with the neighboring parts of the protein and this interaction is stronger than the internal "tension" of structure that do not give this part to fold correctly. And so it was possible to do this in two stages: first to correct this structure alone (ignoring the impact from all other parts), and then we have the correct structure to combine it with other parts of proteins.
Also it would be useful to accelerate the speed of modeling - the smaller elements involved in the simulation, the faster the calculation go. (If the rest of the protein is not simply "frozen", but generally not included in the simulation at all)