[gmx-users] Force field parameters in water or in vacuum?

[Chris Neale]

Interesting, I was not aware of that. For the simulations that I was discussing, we did not see that type of 
collapse on the 10ns timescale, although we started from extended structures and used very high temperatures 
since we were only interested in using MD as one of many methods to generate a massive ensemble of 
conformations that are accessible to the protein and from which we would later select a small population weighted 
ensemble that best-fits experimental data.

>/ The parameters themselves will be the same in water or in vacuum. 
/>/ However, the behaviour of a molecule will be entirely different in water 
/>/ or in vacuum since vacuum lacs the hydrophobic effect that falls out of 
/>/ water being in the system. The hydrophobic effect is, in my opinion, one 
/>/ of the great success stories of MD. On the other hand, proteins will 
/>/ change their structure much faster in vacuum since the friction is much 
/>/ less in that case. So if you are looking to generate structures for some 
/>/ other purpose without caring if the sampling itself is Boltzmann (e.g. 
/>/ generating structures for use with the ENSEMBLE program to characterize 
/>/ intrinsically unfolded proteins by Dr. Julie Forman-Kay) then in vacuuo 
/>/ may be the best way to go.
/>/ 
/Actually in our experience this is not the case. In vacuum the proteins 
collapse and do hardly move away from this structure. See
Biochemistry 46 pp. 933-945 (2007)
JPCB 111 pp. 13147-13150 (2007)
JACS 129 pp. 6726-6735 (2007)
Int. J. Mass Spectrom. 248 pp. 124-135 (2006)