[gmx-users] how to apply constraints in energy minimization

[Sadaf Rani]

Dear Gromacs users

I am trying to run a protein-ligand free energy calculation in which I
initially run MD for 3 ns without free energy calculation and took these
coordinates for free energy in which I am doing two-step energy
minimization.

Firstly at steepest decent with force tolerance of 100 with 20000 nsteps
which converges to in 7992 steps as mentioned below:-
 Energies (kJ/mol)
           Bond Restraint Pot.          Angle    Proper Dih.  Improper Dih.
    1.21424e+03    9.53698e+01    4.72990e+03    1.83362e+04    1.86005e+02
          LJ-14     Coulomb-14        LJ (SR)  Disper. corr.   Coulomb (SR)
    6.81666e+03    7.15551e+04    3.32691e+05   -6.56537e+03   -2.69738e+06
   Coul. recip.    Angle Rest.     Dih. Rest.      Potential Pres. DC (bar)
    4.52484e+03    6.92787e+00    2.04511e+02   -2.26358e+06    0.00000e+00
 Pressure (bar)    dVremain/dl       dEkin/dl      dVcoul/dl       dVvdw/dl
   -2.68131e+04    0.00000e+00    0.00000e+00    6.26922e+03   -1.62832e+03
 dVrestraint/dl   Constr. rmsd
    0.00000e+00    2.65292e-06





*Steepest Descents converged to Fmax < 100 in 7992 stepsPotential Energy  =
-2.2635825e+06Maximum force     =  9.0506676e+01 on atom 16669Norm of force
    =  3.6179803e+00*



    Step           Time
           2470     2470.00000

   Energies (kJ/mol)
           Bond Restraint Pot.          Angle    Proper Dih.  Improper Dih.
    7.44151e+04    9.26252e+01    4.47510e+04    1.83016e+04    1.95586e+02
          LJ-14     Coulomb-14        LJ (SR)  Disper. corr.   Coulomb (SR)
    6.77053e+03    7.14269e+04    4.17453e+05   -6.56537e+03   -3.03191e+06
   Coul. recip.    Angle Rest.     Dih. Rest.      Potential Pres. DC (bar)
    4.48733e+03    6.63779e+00    2.08432e+02   -2.40037e+06    0.00000e+00
 Pressure (bar)    dVremain/dl       dEkin/dl      dVcoul/dl       dVvdw/dl
   -7.77641e+01    0.00000e+00    0.00000e+00    6.49503e+03   -1.88623e+03
 dVrestraint/dl
    0.00000e+00
For the second minimization, I am using l-bfgs integrator at force
tolerance 10 with 50000 nsteps. I get the following message:-



*Energy minimization has stopped, but the forces have not converged to
therequested precision Fmax < 10 (which may not be possible for your
system). *
Double precision normally gives you higher accuracy, but this is often not
needed for preparing to run molecular dynamics.

Low-Memory BFGS Minimizer converged to machine precision in 2471 steps,
but did not reach the requested Fmax < 10.
Potential Energy  = -2.4003668e+06
Maximum force     =  2.1663138e+02 on atom 1093
Norm of force     =  3.8024703e+00

Performed 2565 energy evaluations in total.
But my system crashes at nvt step generating different pdb structures. How
should I solve this
Any suggestions will really be appreciated.

Thanks.

Sadaf Rani
Visiting Ph.D. student
Lancaster University
Uk