VMD-L Mailing List

From: Peter Mawanga (peter.mawanga.lagos_at_gmail.com)
Date: Mon Jun 04 2018 - 10:24:09 CDT

Hello Fotis

You have correctly identified that the issue was primarily with PME and
Martini force-field as such. Also, you have accurately pointed out all the
issues viz.

   1. Change L-J cutoff to 9-12.
   2. Turn "martiniswitching" on.
   3. Keep "PME" off.
   4. Enable "cosangles".
   5. Most importantly, set "dielectric" to 15.

If someone would like to implement this in future, I would strongly
recommend to use a modified version of "namdenergy.tcl" incorporating the
desired parameters and flags and source it via "TkConsole" and run your
"namdenergy" command.

On Wed, May 23, 2018 at 8:07 PM, Fotis Baltoumas <fbaltoumas_at_biol.uoa.gr>
wrote:

> Hello,
>
> By default, NAMDEnergy is tailored for CHARMM/AMBER simulations. However,
> from what I see, you're trying to do calculations with Martini.
>
> Martini uses different switch/cut-off ranges for Lennard-Jones and Coulomb
> interactions, namely, a 9-12 range for Lennard-Jones, and a 0-12 range for
> Coulomb electrostatics. The 0-12 coulomb options, by default, are
> hardcoded in the martiniSwitching option which, to my understanding, is not
> written to the configuration file generated by NAMDEnergy.
>
> Furthermore, martini requires "cosAngles" to be enabled, in order to
> properly model angle potentials.
>
> Finally, unless you are using the polarizable water model and a dielectric
> constant lower than 15, PME is pretty much useless with Martini.
>
> Any of the three aforementioned reasons could be the cause of your crash,
> most likely the lack of 0-12 coulombic switch/cutoff values that influence
> electrostatics. The only reason you didn't see it in your first
> calculation is probably the fact that you used a subset of your system. In
> fact, I would suggest that you re-evaluate the results of the calculation
> that didn't crash, since they are most likely erroneous.
>
> You can do the following: use NAMDEnergy's '-debug' flag so that you keep
> all temporary files produced. Then, open the configuration file that was
> produced (namd-temp.namd) and change all topology/force-field related
> options to the ones you used in your simulation (martiniSwitching on,
> cosAngles on, etc). Make sure that you keep the following lines:
>
> pairInteraction on
> pairInteractionGroup1 1
> pairInteractionFile namd-temp.pdb
> pairInteractionGroup2 2 ;# this is when you want to do calculations with a
> second selection, otherwise comment it out
> #pairInteractionSelf on ;#this is for when you want only one
> selection, if so, uncomment it and comment out pairInteractionGroup2
> coordinates namd-temp.pdb
>
> You can find more about them here: http://www.ks.uiuc.edu/
> Research/namd/2.12/ug/node75.html
> Briefly, the namd-temp.pdb is your system, but with extra labeling for the
> atoms (by default, in their b-factor column) in the selections you use in
> your calculations. (1 for the first selection, 2 for the second selection)
>
> Then, you can run your calculation using NAMD itself:
>
> namd2 namd-temp.namd > energy.log
>
> The produced log file will just have the reported energy terms for each
> timestep. You can then parse it with a script or NAMD plot and draw curves
> or compute averages.
>
> Basically, the above is what NAMDEnergy essentially does: it prepare sa
> pdb file with proper b-factor labels for the selection, it writes a
> configuration file with the options you set in the GUI/tkConsole and with
> pairInteraction terms, it calls namd2 for the calculation and, finally, it
> parses the output log and isolates energy terms.
>
> Good luck,
> Fotis
>
>
>
> On 05/23/2018 04:42 PM, Peter Mawanga wrote:
>
> Greetings to everyone
>
> I calculated the non-bonded interactions (VdW and Electrostatic) between
> two selections, using the same "cutoff" and "switch" distance as employed
> in simulations:
>
> ============================================================
> ================
>
> namdenergy -vdw -elec -nonb -sel [atomselect top "segname AP1"]
> [atomselect top "segname AP2"] -exe ../namd2 -extsys
> relaxation/equil_out.xsc -T 310 -switch 9 -cutoff 12 -pme -par
> ../martini-protein-bonds.par -par ../martini-protein-angles-cos.par -par
> ../martini-protein-dihedrals.par -par ../martini-all-nonb.par -ofile
> "nonb_pep.dat"
>
> ============================================================
> ================
>
>
> However, when I tried to calculate the interaction energy within the
> system:
>
> ============================================================
> ================
>
> namdenergy -elec -sel [atomselect top all] -exe ../namd2 -extsys
> relaxation/equil_out.xsc -T 310 -switch 9 -cutoff 12 -pme -par
> ../martini-protein-bonds.par -par ../martini-protein-angles-cos.par -par
> ../martini-protein-dihedrals.par -par ../martini-all-nonb.par -ofile
> "nonb_sys.dat"
>
> ============================================================
> ================
>
>
> I got the following error:
>
> ============================================================
> ================
>
> ------------- Processor 0 Exiting: Called CmiAbort ------------
> Reason: FATAL ERROR: Low global exclusion count! (266 vs 268) System
> unstable or pairlistdist or cutoff too small.
>
> FATAL ERROR: See http://www.ks.uiuc.edu/Research/namd/bugreport.html
>
> Charm++ fatal error:
> FATAL ERROR: Low global exclusion count! (266 vs 268) System unstable or
> pairlistdist or cutoff too small.
>
> FATAL ERROR: See http://www.ks.uiuc.edu/Research/namd/bugreport.html
>
> ============================================================
> ================
>
> Any insight regarding this error would be highly appreciated.
>
> --
> Cheers
> Peter
>
>
> --
> *******************************************
> Fotis A. Baltoumas
> Bioinformatics Postgraduate Programme
> Section of Cell Biology and Biophysics
> Department of Biology, National & Kapodistrian University of Athens
> Panepistimiopolis, Athens 157 01, GREECE
> --------------------------------------
>
> email : fbaltoumas_at_biol.uoa.grhttp://biophysics.biol.uoa.grhttp://bioinformatics.biol.uoa.gr
> *******************************************
>
> 

-- 
Cheers
Peter