Dear Jim,
Thank you for taking a look into my files.
> On 9 Mar 2018, at 19:42, James Kress <jimkress_58.kressworks.org> wrote:
>
> Hi Ruth,
>
> As I suspected, the SCF is not converging in Orca. This may be due to a
> couple of questionable Orca related issues.
Indeed, Orca is encountering SCF problems
> 1) There is are files from a pre-existing Orca run in the same location as
> your run. The files are being used as we see here:
>
> Checking for AutoStart:
> The File: orc_job.gbw exists
> Trying to determine its content:
> ... Fine, the file contains calculation information
> ... Fine, the calculation information was read
> ... Fine, the file contains a basis set
> ... Fine, the basis set was read
> ... Fine, the file contains a geometry
> ... Fine, the geometry was read
> ... Fine, the file contains a set of orbitals
> ... Fine, the orbitals can be read
> => possible old guess file was deleted
> => GBW file was renamed to GES file
> => GES file is set as startup file
> => Guess is set to MORead
> ... now leaving AutoStart
>
> You need to clean these old files out of the working area for your new
> calculations.
It cannot be because of pre-existing files as a new folder was created for running the simulation, hence I can guarantee you that these files are from this specific run.
> 2) The options displayed as the orca input file are "unusual" and are not
> the ones you specify in &orc
What do you mean by unusual? Am I selecting incompatible settings? I’m new to Orca so I would really appreciate any guidance in its use (despite that their website and manual are very clear and helpful).
When I wrote the original email, I didn’t know that settings from the &orc namelist weren’t going to be transferred to the orc_job.inp file because of the use of a template file, Andreas Goetz then gave me this helpful piece of information.
> If you look at page 165 of the Amber16 manual, you'll see that some of the
> entries you specify are not allowed.
>
> Try this:
>
> In your Sander input file:
>
> &orc
> use_template=1
> /
>
> In orc_job.tpl use the following:
>
> #ORCA input file
> ! PBE def2-SVP def2/JK def2/C
> %elprop
> dipole true
> end
> %scf
> Maxiter=500
> Convergence sloppy
> end
>
> Please let me know if this works OK for you. If not, then please send the
> "ptchrg.xyz" and orca_inpfile.xyz and I'll take a closer look.
In the mean time, before seeing your email, I launched a simulation with the following orc_job.tpl file, and it is working for now, but I shall try what you propose:
#ORCA input file
!PBE Def2-SVP Def2/J Def2-SVP/C SLOPPYSCF
%pal
nprocs 1
end
%method
grid 4
finalgrid 6
end
%scf
maxiter 500
SOSCFStart 0.00033
end
> Some questions:
>
> 1) Why are you using "cbasis”?
I’m using a cbasis to speed-up the convergence when calculating the correlation term (if I understand well, correct me please if I got things wrong).
> 2) Do you really need Orca to calculate and print the dipole moments for
> every QM calculation?
I might use this data but I’m actually not sure yet.. Is it too expensive to print it "just in case"?
> 3) sloppy convergence is dangerous. Medium might be better and would get
> you a much better final density.
I’ve chosen a “sloppy” threshold criterion because I initially wanted to perform some non-equilibrium simulations and it would have never converged if I would have selected a higher accuracy criterion. As I encountered multiple SCF problems, I have decided to first minimise the quantum region and then go for the non-equilibrium simulations. Any simulation protocol suggestions would be much appreciated if you would have any.
All the best!
Ruth
> On 9 Mar 2018, at 09:44, Ruth Helena Tichauer <rhtichau.laas.fr> wrote:
>
> Dear Jim,
>
> Please find the entire orc_job.dat file in attachments.
> Thank you for taking a look at it!
>
> Best regards,
>
> Ruth
>
> <orc_job.dat>
>> On 8 Mar 2018, at 18:03, James Kress <jimkress_58.kressworks.org> wrote:
>>
>> What does the whole Orca output look like?
>>
>> Your SCF is not converging but one would need to look at the complete Orca output file to see what's going on and suggest a solution.
>>
>> Jim
>>
>> -----Original Message-----
>> From: Ruth Helena Tichauer <rhtichau.laas.fr>
>> Sent: Thursday, March 08, 2018 4:55 AM
>> To: AMBER Mailing List <amber.ambermd.org>
>> Subject: [AMBER] QM/MM with Orca: selected input parameters not taken into account
>>
>> Dear Amber users and developers,
>>
>> I’m trying to perform a QM/MM molecular dynamics simulation of a protein complex using the following input files:
>>
>> _min.in:
>> M/MM minimization using Orca
>> &cntrl
>> imin=1,
>> maxcyc=10000, ncyc=9000,
>> ntpr=20, ntwx=20, ntwr=2000, ioutfm=1,
>> cut=12.0,
>> ntb=1,
>> ntr=1, restraint_wt = 5.0,
>> restraintmask = ':170,185,@CA,C,N',
>> ifqnt=1
>> /
>> &qmmm
>> qmmask=':168,169,195-196',
>> qmcharge=-2,
>> qm_theory='EXTERN',
>> qmshake=0,
>> qmcut=17.0,
>> qm_ewald=0,
>> /
>> &orc
>> method='PBE',
>> basis='def2-SVP',
>> cbasis='def2-SVP',
>> jbasis='def2-SVP',
>> num_threads=1,
>> convkey='SLOPPYSCF',
>> dipole=1,
>> maxiter=500,
>> use_template=1,
>> /
>>
>> _orc_job.tpl:
>> #ORCA input file
>> %scf
>> SOSCFStart 0.00033
>> end
>>
>> I’m puzzled since in the mdout file, the chosen parameters are correct:
>> &orc
>> | basis = DEF2-SVP
>> | cbasis = DEF2-SVP
>> | jbasis = DEF2-SVP
>> | method = PBE
>> | convkey = SLOPPYSCF
>> | guess = read
>> | scfconv = -1
>> | grid = 4
>> | finalgrid = 6
>> | maxiter = 500
>> | maxcore = 1024
>> | ntpr = 20
>> | num_threads = 1
>> | debug = 0
>> | dipole = T
>> | use_template = T
>> | /
>>
>> but when I look at the orc_job.dat file, because the simulation crashes with the following error message:
>>
>> ERROR
>> This SCF-wavefunction IS NOT FULLY CONVERGED!
>> You can't use it for properties or numerical calculations !
>> Aborting the run ...
>> Please restart calculation (with larger maxiter/different convergence flags)
>>
>> I see that the SCF stopped after 125 cycles and that the energy threshold is not the one I have selected:
>>
>> *****************************************************
>> * ERROR *
>> * SCF NOT CONVERGED AFTER 125 CYCLES *
>> *****************************************************
>> ---------------
>> SCF CONVERGENCE
>> ---------------
>>
>> Last Energy change ... 3.4826e-05 Tolerance : 1.0000e-08
>> Last MAX-Density change ... 1.7469e-03 Tolerance : 1.0000e-07
>> Last RMS-Density change ... 1.0077e-05 Tolerance : 5.0000e-09
>> Last DIIS Error ... 1.1837e-03 Tolerance : 5.0000e-07
>>
>> **** DENSITY FILE WAS UPDATED (orc_job.scfp.tmp) ****
>> **** ENERGY FILE WAS UPDATED (orc_job.en.tmp) ****
>>
>> --------------------------------------------------------------------
>> WARNING
>> The wavefunction IS NOT YET CONVERGED! It shows however signs of
>> convergence. Therefore the wavefunction will be stored and can be
>> used as input for another calculation.
>> DO NOT USE THIS WAVEFUNCTION FOR ANYHTING ELSE. It is NOT RELIABLE
>>
>> What should I do in order for Orca to take into account the parameters I select with my input files and not the default ones? If I provide the orc_job.tpl file, should I re-define all my parameters in it (and not only the additional ones)?
>>
>> Thank you for any insight into this matter.
>>
>> Sincerely,
>>
>> Ruth
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Mar 12 2018 - 04:00:02 PDT
