On Tue, Dec 18, 2012, Wong, Sergio E. wrote:
>
> I am running amber 10 on a linux cluster (12 cores/node, intel Xeon
> EP X5660, with 24GB of RAM per node). When I run an LMOD calculation
> to mimic an induced fit calculation, I use the following restraint
> mask to restrain residues beyond 6 A of a ligand: restraintmask=':413
> > .6.0 '. All goes well, but if I increase the radius to 10A instead
> of 6A, the error below occurs. Also, very large energies occur on
> the last report on the output file (shown below the error message).
> Any ideas of what might cause the error or any suggestions on how
> to resolve it? The corresponding input file text follows the error
> message.
Did you run the lmodprmtop program to modify your prmtop file (to eliminate
zero L-J terms)? This is necessary for LMOD runs, since the search can
created configurations where atoms get quite near to each other.
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 75 -7.8417E+07 2.2520E+12 1.6224E+14 HG1 4185
>
> BOND = 477.7534 ANGLE = 1713.7025 DIHED = 4111.8065
> VDWAALS = -2986.7942 EEL = -29227.3389 EGB = -4925.0859
> 1-4 VDW = 1476.0553 1-4 EEL = ************* RESTRAINT = 342.4850
> EAMBER = *************
>
> End of LMOD requested XMIN relaxation
> Start of LMOD requested XMIN minimization
>
The fact that you get very large, negative, energies for electrostatics is
quite often a symptom of having zero LJ terms for hydrogens. You could also
check the structures around atom 4185 to see if something odd is happening.
...dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Dec 18 2012 - 05:30:05 PST
