RE: Included Topology (.itp) in gromacs to parameter file for NAMD

From: Vermaas, Joshua (Joshua.Vermaas_at_nrel.gov)
Date: Wed Jul 11 2018 - 18:18:44 CDT

Dumb question: are you specifying a "grocoordfile" or a "coordinates" file? You have a pdb extension being passed to grocoordfile I suspect, which *may* make NAMD unhappy. What combination rule are you using instead of 1? 2 ought to also work, since that is what CHARMM uses.

-Josh

On 2018-07-11 15:37:11-06:00 Ivan wrote:

Thank you again for your response. This is where the error happens. I've searched about the combination rule !=1 too.
Info: USING VERLET I (r-RESPA) MTS SCHEME.
Info: C1 SPLITTING OF LONG RANGE ELECTROSTATICS
Info: PLACING ATOMS IN PATCHES BY HYDROGEN GROUPS
Info: RIGID BONDS TO HYDROGEN : ALL
Info: ERROR TOLERANCE : 1e-08
Info: MAX ITERATIONS : 100
Info: RIGID WATER USING SETTLE ALGORITHM
Info: RANDOM NUMBER SEED 1010
Info: USE HYDROGEN BONDS? NO
Info: Using GROMACS format force field!
Info: GROMACS TOPO FILE 50LA_crystal.top
Info: GROMACS COOR FILE 32x50LA_crystal_autopsf_formatted.pdb
Info: USING GEOMETRIC MEAN TO COMBINE L-J SIGMA PARAMETERS
Combination rule != 1 unsupported in DEFAULTS
------------- Processor 0 Exiting: Called CmiAbort ------------
Reason: unexpected call to exit by user program. Must use CkExit, not exit!
[0] Stack Traceback:
  [0:0] [0x1329155]
  [0:1] [0x13296f5]
I'm running NAMD on a HP Pavilion g6 notebook PC. With Ubuntu 16.04 LTS 64bits, 8Gb RAM, Intel Core i7-3632QM CPU 2.2GHzx8.
It's using the integrated graphic card of Intel, but I think the graphic card has no importance here. If it does, I can look for the drivers for the Radeon graphic card of the laptop.
Best regards Josh, and again thank your for your patience and responding my emails.
Have a nice day.
Ivan

On 11 July 2018 at 19:16, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov<mailto:Joshua.Vermaas_at_nrel.gov>> wrote:
Hi Ivan,

This is a NAMD error at runtime. What does the rest of the log look like, and what sort of hardware are you running NAMD on?

-Josh

On 2018-07-11 03:01:08-06:00 Ivan wrote:

Hello again Josh, I was very thankfull when I read your fast response. It's taken me a while to investigate all the options and advance the project. Now I'm facing new problems and I didn't found any good information among all mailing list and google. I've noticed there are lots of questions about this, but unanswered.
When I run the simulation, this error appears:
Charm++ fatal error:
unexpected call to exit by user program. Must use CkExit, not exit!
I haven't found any exit! on my scripts, any idea why could this be happening?
Best regards,
Ivan

On 3 July 2018 at 20:58, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov<mailto:Joshua.Vermaas_at_nrel.gov>> wrote:
Hi Ivan,

Ick. Based on what is described in the GROMACS documentation, where you effectively "make your own" dihedral potential, I don't recall a similar feature existing in NAMD. If your potentials happen to be sinusoidal, you can use the "extraBonds" interface to define them yourself (http://www.ks.uiuc.edu/Research/namd/2.12/ug/node27.html#SECTION00086400000000000000)<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.ks.uiuc.edu%2FResearch%2Fnamd%2F2.12%2Fug%2Fnode27.html%23SECTION00086400000000000000)&data=02%7C01%7CJoshua.Vermaas%40nrel.gov%7Cb6a364aaf16744029d0e08d5e70cd636%7Ca0f29d7e28cd4f5484427885aee7c080%7C0%7C0%7C636668964685874275&sdata=7UnzWdVQILX3OCnH54yxpDmhJHVmxaGqYq%2FJ9vTQi7E%3D&reserved=0> or write a proper CHARMM-style parameter file for your inputs. Is there a particular reason you don't want to use GROMACS? Or failing that LAMMPS, which I recall also has a similar feature?

-Josh

On 2018-07-03 12:44:14-06:00 Ivan wrote:

Josh, thank you very much for your rapid and nice answer. The tabulated table .xgv is a parameter to mdrun. Any advice or idea would be highly appreciated.
Again, thank you very much for replying.
-Ivan

On 3 July 2018 at 18:51, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov<mailto:Joshua.Vermaas_at_nrel.gov>> wrote:
Hi Ivan,

How is this fed into GROMACS? Is the .xvg fed in during the generation of the .top file, or is it a parameter to mdrun? NAMD should be able to read in a GROMACS .top file (see http://www.ks.uiuc.edu/Research/namd/2.12/ug/node14.html)<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.ks.uiuc.edu%2FResearch%2Fnamd%2F2.12%2Fug%2Fnode14.html)&data=02%7C01%7CJoshua.Vermaas%40nrel.gov%7C4ac64fae8c2a47feba6808d5e114f862%7Ca0f29d7e28cd4f5484427885aee7c080%7C0%7C0%7C636662402546085411&sdata=7h0aezYwY9elYo58rMWPQmCNth4oqh2ONe0CJV5Jl%2BI%3D&reserved=0> if it is the first option, with the caveat that NAMD doesn't parse "#include" statements in a .top file, but not if it is the second.

-Josh

On 2018-07-03 02:46:48-06:00 owner-namd-l_at_ks.uiuc.edu<mailto:owner-namd-l_at_ks.uiuc.edu> wrote:

Thank you very much for all the work you do and keeping the NAMD project alive.
I've recently seen an article about a customized force field for GROMACS with an .itp extension and a .xvg tabulated table for the dihedral interactions. I would like to convert that to use it on NAMD and continue to make accurate simulations using NAMD, so it has to have AMBER or CHARMM format I guess.
I've searched the internet and on this site I found this:
"... the Gromacs force field uses a different type of function
(for proper and improper dihedrals) than the Charmm force field. Gromacs
appears to use harmonic and periodic functions while Charmm uses modified
Fourier series. "
http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2004-2005/0968.html%7Cf9b7ce58097847667c6408d5e0c180df%7Ca0f29d7e28cd4f5484427885aee7c080%7C0%7C1%7C636662044069868686&sdata=yygV8kAqLOTJjN5xRCbYQ6rNl5x%2FCFHgsIt2oH8xyFQ%3D&reserved=0>
How can I solve this? Any idea? Is this even possible to do?
Thank you in advance,
Ivan

This archive was generated by hypermail 2.1.6 : Tue Dec 10 2019 - 23:19:59 CST