[gmx-users] How to avoid the infinite potential energy in simulations of dimers?

Qing Liu 15110700076 at fudan.edu.cn
Fri Feb 21 19:01:43 CET 2020


Dear Justin Lemkul,

 Thanks for your reply. I just downloaded the Cryo-EM structure of the dimer from RCSB, added missing residues using Modeller and built systems by the following commands:

    

     gmx pdb2gmx -f dimer.pdb -o dimer.gro -p dimer.top -water tip3p -ff amber99sb-ildn -ignh
      gmx editconf -f dimer.gro -o vac-min-pbc.gro -bt cubic -d 1.5  -c
      gmx solvate -cp vac-min-pbc.gro -cs spc216.gro -p dimer.top -o vac-min-pbc-solv.gro
      gmx grompp -v -f sol-min.mdp -c vac-min-pbc-solv.gro -p dimer.top -o vac-min-pbc-solv.tpr
      gmx genion -s vac-min-pbc-solv.tpr -o vac-min-pbc-solv-salt.gro -conc 0.15 -neutral -pname NA -nname CL  -p dimer.top
      gmx grompp -f sol-min.mdp -c vac-min-pbc-solv-salt.gro -p dimer.top -o vac-min-pbc-solv-salt-min.tpr
      gmx  mdrun -v -deffnm vac-min-pbc-solv-salt-min
      gmx grompp -f pr-md.mdp -c vac-min-pbc-solv-salt-min.gro -p dimer.top -o pr-md.tpr  -r vac-min-pbc-solv-salt-min.gro
      gmx mdrun -nb gpu -v -deffnm pr-md


The contents of sol-min.mdp file are:


      ; Define can be used to control processes
      define          = -DFLEXIBLE
      ; Parameters describing what to do, when to stop and what to save
       integrator      = steep         ; Algorithm (steep = steepest descent minimization)
       emtol           = 1.0           ; Stop minimization when the maximum force < 1.0 kJ/mol
       nsteps          = 5000          ; Maximum number of (minimization) steps to perform
       nstenergy       = 1             ; Write energies to disk every nstenergy steps
       energygrps      = System        ; Which energy group(s) to write to disk
      ; Parameters describing how to find the neighbors of each atom and how to calculate the interactions
     ns_type         = grid          ; Method to determine neighbor list (simple, grid)
     coulombtype     = cut-off       ; Treatment of long range electrostatic interactions
     rcoulomb        = 1.0           ; long range electrostatic cut-off
     rvdw            = 1.0           ; long range Van der Waals cut-off
     constraints     = none          ; Bond types to replace by constraints
     pbc             = xyz           ; Periodic Boundary Conditions (yes/no)
     sc-coul         = yes



The log file of energy minimization shows:
      Steepest Descents converged to machine precision in 2108 steps,
      but did not reach the requested Fmax < 1.
      Potential Energy  = -2.0181694e+07
      Maximum force     =  3.4778253e+02 on atom 114839
       Norm of force     =  3.9966155e+01


The contents of pr-md.mdp are:
     title           = PR MD
    define          =-DPOSRES
    ;run control
    integrator      =md
    tinit           =0
   dt              =0.002
    nsteps          =500000
    comm_mode       =Linear
    nstcomm         =10
   comm_grps       =System
   ;output control
   nstxout         =0
  nstvout         =0
  nstlog          =50000
   nstcalcenergy   =1
   nstenergy       =50000
   nstxtcout       =50000
  xtc_grps        =System
  energygrps      =System
 ;neighbor searching
  nstlist         =10
  ns_type         =grid
  pbc             =xyz
  rlist           =1.4
 ;electrostatics
  coulombtype     =PME
  rcoulomb        =1.4
  ;vdw
  vdwtype         =Cut-off
  rvdw            =1.4
  dispCorr        =EnerPres
  ;Ewald
  fourierspacing  =0.1
  pme_order       =4
  ewald_rtol      =1e-5
 ;temperature coupling
  tcoupl          =v-rescale
  tc_grps         =System
  tau_t           =0.1
  ref_t           =200
  ;velocity generation
  gen-vel         =no
  ;bonds
  constraints     =all-bonds
  constraint_algorithm    =SHAKE
  shake-tol       =0.0001
  morse                   =no
  continuation            =yes
  sc-coul         =yes



The pr-md.log shows:

   Fatal error:
   Step 2460: The total potential energy is nan, which is not finite. The LJ and
   electrostatic contributions to the energy are 3.25168e+06 and -1.99067e+07,
   respectively. A non-finite potential energy can be caused by overlapping
   interactions in bonded interactions or very large or Nan coordinate values.
   Usually this is caused by a badly- or non-equilibrated initial configuration,
   incorrect interactions or parameters in the topology.








--


Best wishes, 

------------------------------------------------------------

Qing Liu

Fudan Univ.

Mobile: +86—13358129621

E-mail: 15110700076 at fudan.edu.cn






>Message: 5
>Date: Fri, 21 Feb 2020 09:21:34 -0500
>From: Justin Lemkul <jalemkul at vt.edu>
>To: gmx-users at gromacs.org
>Subject: Re: [gmx-users] How to avoid the infinite potential energy in
>	simulations of dimers?
>Message-ID: <b1dbc8bb-8640-5834-dd59-9a69a98e591a at vt.edu>
>Content-Type: text/plain; charset=gbk; format=flowed
>
>
>
>On 2/21/20 5:11 AM, Qing Liu wrote:
>> Dear Gromacs users,
>>        I run some simulations of dimers, a protein binding another protein. When these simulations step into equilibriumstages,  they will crash with the following:
>>
>>
>>
>>
>>
>>                    Step 2029: The total potential energy is nan, which is not finite. The LJ and
>>                    electrostatic contributions to the energy are 746119 and -5.71229e+06,
>>                    respectively. A non-finite potential energy can be caused by overlapping
>>                   interactions in bonded interactions or very large or Nan coordinate values.
>>                  Usually this is caused by a badly- or non-equilibrated initial configuration,
>>                   incorrect interactions or parameters in the topology.
>>         
>>
>>        I try to turn on soft-core potential but it's not working. However, the simulations of monomers with same mdp files are normal.  How should I do to solve the problem?
>>
>
>Your initial configuration is unreasonable as you have a massive LJ 
>repulsion. How did you construct the coordinates of the system? Likely 
>energy minimization will have reported unreasonable forces, as well.
>
>-Justin
>
>-- 
>==================================================
>
>Justin A. Lemkul, Ph.D.
>Assistant Professor
>Office: 301 Fralin Hall
>Lab: 303 Engel Hall
>
>Virginia Tech Department of Biochemistry
>340 West Campus Dr.
>Blacksburg, VA 24061
>
>jalemkul at vt.edu | (540) 231-3129
>http://www.thelemkullab.com
>
>==================================================


More information about the gromacs.org_gmx-users mailing list