I carried out an MD run with 'belly
option' turned on for the first 100 ps and then I switched it off.
Then I performed superposition of peptide molecules in consecutive
time frames to find (unexpectedly) that 'frozen atoms' are not
frozen at al. The atom fluctuations are smaller but inner distances
in 'frozen' peptide changes. What's the reason?
If you were running with a box & constant pressure, the box
was being rescaled to provide pressure corrections. When this
happens, all the atomic coordinates are also scaled. When
watching a movie of the solute during initial constant pressure
belly equilibration, you can see the molecule smoothly shrink
toward the origin. For atoms that are allowed to move, the
force field &/or SHAKE restores reasonable geometry after each
box rescale.
Bill Ross
Received on Wed Feb 21 2001 - 10:37:03 PST
