[gmx-users] simulated annealing

[Justin Lemkul]

On 2/19/18 2:31 PM, Ahmed Mashaly wrote:
> Thanks, Justin.
>
> I had set it to 300K before (as the original mdp was), but now I found others with 0
>
>
> Now when if we look at the log file, you can see ref temp for both groups starts from 0 and increasing, but the temperature in the details goes from 300 to 5 to 2, then it starts increasing again:
>
> ----------------------------------------------------------------------
> Started mdrun on rank 0 Thu Feb 15 06:00:49 2018
> Step Time
> 0 0.00000
>
> Current ref_t for group Protein: 0.0
> Current ref_t for group non-Protein: 0.0
> Energies (kJ/mol)
> Bond Angle Proper Dih. Improper Dih. LJ-14
> 3.51283e+03 1.20582e+04 6.93519e+04 6.27988e+02 1.88294e+04
> Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) Coul. recip.
> 2.26883e+05 7.62987e+05 -3.36842e+04 -5.35584e+06 1.97969e+04
> Potential Kinetic En. Total Energy Conserved En. Temperature
> -4.27548e+06 7.07199e+05 -3.56828e+06 -3.56828e+06 3.00605e+02
> Pres. DC (bar) Pressure (bar) Constr. rmsd
> -1.86887e+02 1.45163e+03 4.71563e-06
>
> DD step 19 load imb.: force 93.3% pme mesh/force 1.284
>
>
> step 40 Turning on dynamic load balancing, because the performance loss due to load imbalance is 3.6 %.
>
> DD load balancing is limited by minimum cell size in dimension Y
> DD step 499 vol min/aver 0.501! load imb.: force 6.8% pme mesh/force 0.849
>
> Step Time
> 500 1.00000
>
> Current ref_t for group Protein: 0.6
> Current ref_t for group non-Protein: 0.6
> Energies (kJ/mol)
> Bond Angle Proper Dih. Improper Dih. LJ-14
> 3.46055e+03 1.23246e+04 6.95866e+04 6.62756e+02 1.87622e+04
> Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) Coul. recip.
> 2.26321e+05 7.48416e+05 -3.36842e+04 -5.50170e+06 1.60271e+04
> Potential Kinetic En. Total Energy Conserved En. Temperature
> -4.43983e+06 1.20712e+04 -4.42776e+06 -3.56007e+06 5.13104e+00
> Pres. DC (bar) Pressure (bar) Constr. rmsd
> -1.86887e+02 -1.52962e+03 2.60669e-06
>
>
>
> DD load balancing is limited by minimum cell size in dimension Y
> DD step 999 vol min/aver 0.536! load imb.: force 4.4% pme mesh/force 0.719
>
> Step Time
> 1000 2.00000
>
> Current ref_t for group Protein: 1.2
> Current ref_t for group non-Protein: 1.2
> Energies (kJ/mol)
> Bond Angle Proper Dih. Improper Dih. LJ-14
> 3.36670e+03 1.22399e+04 6.95391e+04 6.50404e+02 1.87726e+04
> Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) Coul. recip.
> 2.26254e+05 7.67224e+05 -3.36842e+04 -5.55850e+06 1.52291e+04
> Potential Kinetic En. Total Energy Conserved En. Temperature
> -4.47891e+06 5.32623e+03 -4.47359e+06 -3.56020e+06 2.26399e+00
> Pres. DC (bar) Pressure (bar) Constr. rmsd
> -1.86887e+02 -1.51090e+03 2.52820e-06
>
> DD load balancing is limited by minimum cell size in dimension Y
> DD step 1499 vol min/aver 0.568! load imb.: force 3.1% pme mesh/force 0.727
>
> Step Time
> 1500 3.00000
>
> Current ref_t for group Protein: 1.8
> Current ref_t for group non-Protein: 1.8
> Energies (kJ/mol)
> Bond Angle Proper Dih. Improper Dih. LJ-14
> 3.30312e+03 1.22047e+04 6.95879e+04 6.55239e+02 1.88651e+04
> Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) Coul. recip.
> 2.26345e+05 7.77640e+05 -3.36842e+04 -5.58600e+06 1.46894e+04
> Potential Kinetic En. Total Energy Conserved En. Temperature
> -4.49639e+06 5.57315e+03 -4.49082e+06 -3.56022e+06 2.36895e+00
> Pres. DC (bar) Pressure (bar) Constr. rmsd
> -1.86887e+02 -1.43160e+03 2.50109e-06
> ---------------------------------------------------------------------
>
> And when I used gmx energy to generate temp curve, I got this at the beginning:
>
> 0.000000 300.605164 300.543854 300.611908
> 1.000000 5.131036 3.108571 5.354491
> 2.000000 2.263988 1.550216 2.342851
> 3.000000 2.368947 1.869242 2.424157
> 4.000000 2.709096 2.333785 2.750562
> 5.000000 3.236674 3.030164 3.259491
> 6.000000 3.733118 3.523592 3.756267
> 7.000000 4.282474 3.992255 4.314539
> 8.000000 4.829560 4.523222 4.863406
> 9.000000 5.418922 5.164090 5.447078
>
>
> Also I noticed a shrink in the water box after NVT, and it expands to the usual at the start of NPT, is the shrink realted to this?

Possibly. If you tell the system it's at 300 K, but then force the 
thermostat to try to damp the velocities down to zero, you're going to 
get weird physical behavior.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.biochem.vt.edu/people/faculty/JustinLemkul.html

==================================================