A puzzled biologist here! Is molecular dynamics simulations actually run on a single protein molecule to give the structure with minimum Gibbs free energy or does it consider many molecules at once (what I think ensemble is meant to mean)?
4
-
$\begingroup$ either I do not understand the question or the answer that you have accepted. Usually an MD simulation is run in a cell which contains a molecule (say protein) and a solvent (mostly water). To be able to see the protein molecule clearly usually the results presented by removing the surrounding solvent. $\endgroup$– physicopathCommented May 5, 2017 at 9:21
-
1$\begingroup$ see for example pnas.org/content/99/10/6719.full.pdf. In the second paragraph of the paper they say that they have water but in the figures you never see the water molecules $\endgroup$– physicopathCommented May 5, 2017 at 9:24
-
$\begingroup$ Thank you for the point and for the article. In my question, by "single protein molecule" I meant 1 box containing 1 protein and surronding water molecules, vs having many such boxes, and what I got from the answer was that , for ergodic systems, as the time passes and there are "different conformations" of a single protein molecule in 1 box, it is as though we have a snapshot(I am not sure about this "snapshot") of "different conformations" in many boxes. $\endgroup$– biofanCommented May 6, 2017 at 14:36
-
$\begingroup$ OK. Keep in mind also that usually periodic boundary conditions are applied in MD simulations, that is, you see one box (super cell) but that box is repeated in three dimensions $\endgroup$– physicopathCommented May 6, 2017 at 16:08
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Any given MD simulation acts on only a single instance of the molecule. It simply integrates Newton's second law to produce the trajectory of that molecule over time.
The ensemble refers to the theoretical class of every possible configuration that satisfies some constraint (it's a concept from statistical mechanics). In practice we make the assumption (the ergodic hypothesis) that if you average a quantity over a trajectory for long enough then you will get the same answer as the ensemble average.
