Hi Tamara,
2. For the template file to be read, 'use_template = 1' needs to be
specified in the &gau input block.
1. For the QM/MM cutoff being too large, the same rules apply as for the
classical electrostatic cutoff; it can't be larger than half the box
length, otherwise you'll have overlapping images. Right now you can see
that the largest sphere that will fit inside the box is 34.633 Angstroms,
but you can't actually set your cutoff to that value since the size of the
QM region needs to be taken into account. It's something you'll have to
calculate or play around with.
As for the "Error with system call (executing Gaussian)", a quick peek at
the code shows that Amber is having trouble calling Gaussian with the input
file. Are you sure that it's in your path and not aliased in a strange way?
A separate note about your template file; you're calling 'opt', but you
don't need to, since Amber takes the forces that Gaussian calculates and
uses those in its internal minimizer. The geometry optimization is taken
care of completely by Amber.
Hope this helps.
Eric
On Mon, Jun 17, 2013 at 8:39 PM, Tamara Meirelles <
tamara.meirelles.anu.edu.au> wrote:
> Hi,
>
> I want to use US-QM/MM simulations to determine the activation barrier of
> an enzyme catalyzing a transformation of its substrate. I have heated and
> equilibrated the system classically and now I want to do a QM/MM
> minimization and equilibration before a production run with umbrella
> sampling. I am using AMBER12 and Gaussian09.
>
> My minimization input file looks like this:
>
> minimization system
> &cntrl
> imin = 1,
> maxcyc=2000, ncyc = 500,
> ntpr = 10, ntwr = 1000,
> ntf = 1, ntc = 1, cut = 8.0,
> ntb = 1, ntp = 0,
> ifqnt = 1,
> nmropt = 1, ioutfm = 1,
> &end
> &wt
> type='END',
> &end
> DISANG=disang
> &qmmm
> qmmask =
> '.1-35,2139-2140,2152-2165,3408-3409,3419-3435,5509-5514,7431-7441'
> qmcharge = -1,
> qm_theory = 'EXTERN',
> qm_ewald = 0,
> qmmm_switch = 1, r_switch_hi = 8.0, r_switch_lo = 6.0,
> qmmm_int = 2,
> qmshake = 0,
> printdipole = 1,
> writepdb = 1,
> qmcut = 500.0,
> scfconv = 1.0d-9,
> diag_routine = 0,
> peptide_corr = 1,
> /
> &gau
> num_threads=8,
> charge = -1,
> /
>
> The disang file is as follows:
>
> Harmonic distance restraints
> &rst
> iat=18,3428,
> r1=-0.0, r2=1.7, r3=1.7, r4=90.0,
> rk2=200.0, rk3=200.0,
> /
>
> And the gau_job.tpl:
>
> #P wB97XD/6-31+G(d,p) OPT nosymm
>
> qmmm
>
> -1 1
>
> When I run this, it aborts with the following message (I have also attached
> the full output file):
>
> ****************************************************
> ERROR: QM region + cutoff larger than box dimension:
> QM-MM Cutoff = 500.0000
> Coord Lower Upper Size Radius of largest sphere inside unit
> cell
> X -504.740 503.106 1007.846 34.633
> Y -507.088 508.633 1015.721 34.633
> Z -508.134 510.109 1018.243 34.633
> ****************************************************
> SANDER BOMB in subroutine QM_CHECK_PERIODIC<qm_mm.f>
> QM region + cutoff larger than box
> cannot continue, need larger box.
>
> Setting qmcut to around 100A has the same effect. If I set qmcut to a lower
> value, e.g. 15 A, it crashes with the following message:
>
> SANDER BOMB in subroutine get_gau_forces (qm2_extern_gau_module)
> Error with system call (executing Gaussian)
> Will quit now.
>
> Which I assume is due to the cutoff not being large enough (?)
> It also caught my attention that the gau_job.inp file produced in this run
> looks like:
>
> %chk=gau_job.chk
> %NProcShared=8
> #P BLYP/6-31G* SCF=(Conver=8) Force Punch=Derivatives NoSymm Charge
> Prop=(Field,Read)
>
> Gaussian run using SANDER external interface.
>
> -1 1
>
> which means that it did not read the functional and basis set I gave it,
> although it did read the number of processors to use from the AMBER input
> file.
>
> So, to sum up, my questions are:
> 1) What should I do in order to get a cutoff large enough but that at the
> same time does not cause this sander bomb?
> 2) Why is Gaussian not reading the .tpl file?
>
> Thank you in advance for your help.
>
> Tamara
>
> --
> *P *Please consider the environment before printing my email.
>
> _______________________________________________
> 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 Wed Jun 19 2013 - 07:30:03 PDT
