Re: error when doing minimization with the charmm Drude FF

From: Frank Haverkort (
Date: Fri Aug 30 2013 - 08:48:43 CDT

Dear NAMD users,

As you could read in my previous mail, I'm trying to perform an energy
minimization with NAMD using the Charmm drude force field. I've been
able to reproduce the error in my previous mail using only files from
the NAMD website (except for the config file), which makes me wonder if
the problem is caused by NAMD rather than my input files.

I took the sample pdb, psf and inp files from (section Sample simulation files,
water (216 molecules)), and am using NAMD version
NAMD_2.9_Linux-x86_64-multicore. I still get exactly the same error message:

If anyone has an idea what is going on that would be great, but even if
you don't, could someone please tell me whether it is at all possible to
do an energy minimization in NAMD on a system that contains lone pairs
(like SWM4-NDP water). After reading the manual I was convinced this
should be possible, but now I'm not so sure anymore.

Thanks very much for your time!

Op 28-08-13 12:23, Frank Haverkort schreef:
> Hi,
> I'm trying to minimize a box of dye molecules in water using the
> Charmm Drude force field. However, I get the error:
> Also, in the "structure summary" section of NAMD's standard output,
> the stated number of bonds in the system is too low by exactly the
> number of bonds to lone pairs (these lone pairs belong to the SWM4-NDP
> water). It thus seems that bonds to lone pairs are not recognized by
> To solve this problem, I tried changing "drude off" into "drude on" in
> the configuration file, since I think lone pairs are part of the
> charmm Drude FF but not of the nonpolarized charmm one. However, I
> then got an error message that stated I needed a Langevin thermostat,
> and when I also added lines for a Langevin thermostat in the
> configuration file, I got a new error message stating you can't
> minimize and use the langevin thermostat simultaneously. So this
> turned out to be a dead end.
> In another attempt at trying to figure out what was wrong, I ran a
> normal MD simulation with exactly the same psf and pdb files as in the
> minimization (with very small time step of 0.1 fs to account for the
> missing minimization). This simulation ran without problems, so my psf
> and pdb files are apparently OK.
> I have no clue what else to try, and would appreciate any possible help!
> The best,
> Frank
> This is the configuration file that I use:
> minimization on
> set boxsize 60.0
> #input files
> coordinates box_all.pdb
> structure box_all_with_polarizability.psf
> parameters charmm_drude_c36_aug12.prm
> paraTypeCharmm on
> #output file (section 3.2.2)
> outputname out #specifies the prefix for the output files
> binaryoutput no #activates binary output (yes=binary)
> outputenergies 100 #number of timesteps between energy output
> # integrator params
> firsttimestep 0 # Starting time of the simulation.
> rigidBonds water
> useSettle on
> # force field params
> seed 12345
> waterModel swm4
> exclude 1-3
> switching on # Use a switching function to truncate the van der
> Waals potential energy smoothly at the cutoff distance.
> switchdist 10.5
> cutoff 12.0 # For van der Waals and electrostatic forces.
> pairlistdist 13.5
> fullElectFrequency 4 # Number of timesteps between evalutations of
> long-range electrostatic forces. MODIFY WHEN TIMESTEP LARGER THAN 1 FS
> longSplitting c2 # Method to split electrostatic forces in short range
> and longe range part.
> # drude
> drude off
> # periodic cell
> cellBasisVector1 $boxsize 0 0
> cellBasisVector2 0 $boxsize 0
> cellBasisVector3 0 0 $boxsize
> # long-range electrostatics
> PME yes
> PMEGridSpacing 1.0
> FFTWUseWisdom no # This can be turned on later, when using the same
> size PME grid multiple times, to reduce startup time.
> minimize 10000

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:23:38 CST