AMBER Archive (2003)

Subject: RE: question on EWALD

From: Yong Duan (yduan_at_udel.edu)
Date: Mon Feb 17 2003 - 16:54:14 CST


Dear Sichun,

You may or may not know that the density is often at around 0.8g/cc
during initial model building (for typical 10K atom systems). This
happens because one needs to remove the water molecules around the
solute and needs to give additional cushion to avoid close-contacts
across the unit cells. These are described in the manual. Of course,
this typical 0.8g/cc is not really the state that you try to simulate.
To reach a reasonable density (around 1g/cc), one needs to conduct a
relatively short simulation at room temperature with NPT. This is the
justification. I am not sure what do you mean by physics.

yong
*****

I need to make it clear for myself. Do you mean using NPT to get a
proper density instead of NVT? And what is the physics behind?

Thanks again,
Sichun