Re: pair interactions in NPT

From: Philip Fowler (
Date: Fri Sep 02 2005 - 02:22:21 CDT


The way to do this is to include the unitcell info in the DCD using
DCDUnitCell on in the NAMD input file (I believe that this on by the
default for NAMD2.5 and later). Then when you read the DCD back into
NAMD to calculate the pair interaction energy NAMD retrieves the
CellBasisVectors from the DCD file itself and correctly wraps the
water etc. This would be pretty obvious if it got it wrong as it is
likely NAMD would at some point try wrapping a water on top of
another water and infinite energies would start to appear.

If you ran multiple simulations and then concatenated the DCDs using
catdcd, then your problem may be that the catdcd supplied with NAMD
strips out the unitcell info from the
DCD files. If this is the case you will need to recompile catdcd as
described in the thread below.


> Hi,
> I need to calculate pair interaction energy after a simulation in
> NPT. For each sets of coordinates from a dcd file I have different
> sets of CellBasisVectors. I took the script below and I adopted it
> in the way that it changes not only the fistTimestep but also the
> CellVectors. It didn't work!
> After several modifications I discovered that NAMD can not change
> vectors values during a simulation (it stops with information
> "FATAL ERROR: Setting parameter cellBasisVector1 from script failed").
> I absolutely need to calculate the interaction energy for my
> simulations, what can I do?
> Any help will be very appreciated!
> Anna
> "coorfile open dcd name.dcd
> set ts 0
> while { [coorfile read] != -1 } {
> incr ts
> firstTimestep $ts
> run 0
> }
> coorfile close"

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:06 CST