Hi Ross,
I merely tried to say that the "best" choice of parameters may be
system-specific.
The system was equilibrated for 1ns (see the initial inpcrd and mdin
attached). OTOH, the energy must be conserved in NVE irrespectively of the
initial configuration. OTOH2: it is well-known that Verlet
(with any timestep > 0) does not preserve energy (the original
Hamiltonian).
Have a great day,
Volodymyr
On Wed, September 10, 2008 08:22, Ross Walker wrote:
> Hi Volodymyr,
>
> So dsum_tol of 10-6 improves things here but I am shocked at how bad the
energy conservation is here. Was this system properly equilibrated? Is
the
> density good for the starting coordinates. I know 2fs time step can
sometimes affect energy conservation but it generally isn't that bad.
>
> All the best
> Ross
>
>> -----Original Message-----
>> From: owner-amber-developers.scripps.edu [mailto:owner-amber-
>> developers.scripps.edu] On Behalf Of Volodymyr Babin
>> Sent: Monday, September 08, 2008 10:48 AM
>> To: amber-developers.scripps.edu
>> Subject: amber-developers: dsum_tol and shake tolerance on energy
conservation (another example)
>> Dear All,
>> please see another example that might be of interest
>> (ethanol using gaff) attached.
>> With kind regards,
>> Volodymyr
>
>
- application/octet-stream attachment: mdin.0
Received on Thu Sep 11 2008 - 08:41:59 PDT
