AMBER Archive (2005)

Subject: Re: AMBER: QMMM question

From: Guanglei Cui (cuigl_at_csb.sunysb.edu)
Date: Mon Oct 10 2005 - 09:14:20 CDT


Hi,

 From my limited experience, in sander.QMMM the QM regions are specified
by a namelist variable "labels" in &cntrl. You have something different
in your miniQMMM.in.in and mdQMMM.in. Perhaps they're equivalent. Could
that be the reason of the error?

Out of curiosity, you had quite a few restraints defined? Which one is
your reaction coordinates?

Guanglei

xiaoqin huang wrote:
> Dear amber users, I have a question on QMMM with Amber8.
> I want to do the umbrella sampling for the reaction: CL (ion, -1e) +
> CH3CL-----CLCH3 + CL(ion, -1e), and calculate the potential of mean
> force (PMF) of this reaction, I solvatecap the solute (the whole
> reactants, CL +CH3CL) at 18A size, and add one counterion (Na+) to
> balance the charge of whole system. All reactant atoms (totally 6, not
> include the Na+) are treated by QM, while all others are as MM.
> I starte relaxing the system for 5000 steps, and it run smoothly over
> and get the restart file (input file miniQMMM.in is attached)
> Next, I want to do the MD for the whole system, I use the NMR
> constraints for the reaction coordiamtes sampling, the restraint file is
> dist.RST. however, after the input file is read (attached mdQMMM.in is
> the input file), it stopped with a fatal error as:
> -----------------------------------------------------------------------------
>
> Number of triangulated 3-point waters found: 724
> FATAL ERROR
> Link atoms MUST be optimized before MD run
> ----------------------------------------------------------------------------
>
>
> by the way, I have the divcon.in (assign the charge of QM part of -1e)
> under the subdirectory where the job is submitted.
>
> I have compiled the amber8 in serial, and patch the bugfix.all, but get
> some rejects (attached file rjects).
>
> is there anyone could help me find what is wrong, whether the input file
> or something else?
>
> I attached the input files both for minimization and MD, please have a
> look.
>
> thanks very much!
>
> xiaoqin huang
>
> college of pharmacy
> university of kentucky
>
> ***************
> *** 792,803 ****
>
> call fill_xtran(xtran)
> call fill_tranvec()
> - call get_stack(l_tau,mxeedtab)
> - call fill_eed_table(eedtbdns,mxeedtab, &
> - x(leed_cub),r_stack(l_tau),x(leed_lin),eedmeth,ee_type)
> - call free_stack(l_tau)
> call init_coulomb_switch(cutoffnb,dxdr, &
> - x(leed_cub),x(leed_lin),eedtbdns,eedmeth,ee_type)
> call vdw_correct_setup(numatoms,ix(i04),ntypes,ix(invdwcls))
> if( icfe /= 0 ) &
> call vdw_correct_setup(numatoms,ix(i84),ntypes,ix(invdwcls2))
> --- 792,805 ----
>
> call fill_xtran(xtran)
> call fill_tranvec()
> + if (periodic /= 0) then !Do not need this memory for non-periodic sim.
> + call get_stack(l_tau,mxeedtab)
> + call fill_eed_table(eedtbdns,mxeedtab, &
> + x(leed_cub),r_stack(l_tau),x(leed_lin),eedmeth,ee_type)
> + call free_stack(l_tau)
> + end if
> call init_coulomb_switch(cutoffnb,dxdr, &
> + x(leed_cub),x(leed_lin),eedtbdns,eedmeth,ee_type)
> call vdw_correct_setup(numatoms,ix(i04),ntypes,ix(invdwcls))
> if( icfe /= 0 ) &
> call vdw_correct_setup(numatoms,ix(i84),ntypes,ix(invdwcls2))
> ***************
> *** 810,818 ****
> call mpi_comm_size(recip_comm,numtasks,ierr)
> call mpi_comm_rank(recip_comm,mytaskid,ierr)
> #endif
> - call pmesh_kspace_setup( &
> - x(lprefac1),x(lprefac2),x(lprefac3),x(lfftable), &
> - nfft1,nfft2,nfft3,order,sizfftab,sizffwrk,opt_infl,ew_type)
> #ifdef MPI
> call mpi_comm_size(world_comm,numtasks,ierr)
> call mpi_comm_rank(world_comm,mytaskid,ierr)
> --- 812,822 ----
> call mpi_comm_size(recip_comm,numtasks,ierr)
> call mpi_comm_rank(recip_comm,mytaskid,ierr)
> #endif
> + if (periodic /= 0) then !do not need to do this for non-periodi
> + call pmesh_kspace_setup( &
> + x(lprefac1),x(lprefac2),x(lprefac3),x(lfftable), &
> + nfft1,nfft2,nfft3,order,sizfftab,sizffwrk,opt_infl,ew_type)
> + end if
> #ifdef MPI
> call mpi_comm_size(world_comm,numtasks,ierr)
> call mpi_comm_rank(world_comm,mytaskid,ierr)
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu