From: Thomas Evangelidis (tevang3_at_gmail.com)
Date: Sat Dec 01 2012 - 12:06:55 CST
I should add that the difference in VDW isn't really a round-off error ---
it's still with the output/reporting system --- my code wasn't as formatted
as the regular NAMD code so the output has different number of significant
digits (another thing to fix for me). If aMD total boost is used, the VDW
used in aMD will be exactly the same as the system's VDW.
Yi
On 1 December 2012 20:03, Thomas Evangelidis <tevang3_at_gmail.com> wrote:
> Hi Thomas,
> I didn't realize that you were using Implicit solvent --- this is something new
> in NAMD (the theory is around for a long time but only very recently it was
> added to NAMD). If you look at the VDW energy, they're actually very close ---
> within round off errors. The only glaring difference is ELECT. I didn't check
> the code yet but I'm almost certain that this difference is purely caused by aMD
> reporting the wrong types of the electrostatic energy --- the PME term is not
> calculated in implicit solvent simulation but this term is hard-coded into our
> aMD energy reporting formula. I think I should have more time to check more
> carefully this weekend and either solve the discrepancy or put a note on NAMD
> manual.
>
> ps, before it's fixed, this would mean that aMDT (boosting total potential) is
> incompatible with implicit solvent simulation.
>
> yi
>
> On Nov 30, 2012, at 12:22 AM, Thomas Evangelidis wrote:
> > Hi Yi,> > Yes, output 1 and 3 are from the same .log file, so are outputs 2 and 4.> That's why I asked you what to trust, the "ACCELERATED MD:" or the> "ENERGY:" line?> > For your convenience I have attached the whole log files. The NAMD version> was CVS-2012-11-19.> > best,> Thomas
>
>
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:22:19 CST