Изучение работы методов контроля температуры в молекулярной динамике

В этом задании нужно было изучить, как работает контроль температуры в молекулярной динамике на примере GROMACS.

Для примера выбрана молекула этана.

Скачаем все необходимые файлы для выполнения задания.

%%bash
wget http://kodomo.fbb.msu.ru/FBB/year_08/term6/be.mdp
wget http://kodomo.fbb.msu.ru/FBB/year_08/term6/vr.mdp
wget http://kodomo.fbb.msu.ru/FBB/year_08/term6/nh.mdp
wget http://kodomo.fbb.msu.ru/FBB/year_08/term6/an.mdp
wget http://kodomo.fbb.msu.ru/FBB/year_08/term6/sd.mdp
wget http://kodomo.fbb.msu.ru/FBB/year_08/term6/etane.gro

# загрузим модули
import rdkit
from rdkit import Chem
from rdkit.Chem import AllChem, Draw
from rdkit.Chem import Descriptors
from rdkit import RDConfig
from IPython.display import Image,display
import numpy as np

# создадим этан
mol=Chem.MolFromSmiles('CC')
AllChem.Compute2DCoords(mol)
m3d=Chem.AddHs(mol)
Chem.AllChem.EmbedMolecule(m3d)
AllChem.MMFFOptimizeMolecule(m3d,maxIters=500,nonBondedThresh=200 )

Нам нужно создать файл топологии et.top для этана. Будем создавать его по этапам.

##
et_file = open('et.top', 'w')
head = '''#include "/usr/share/gromacs/top/oplsaa.ff/forcefield.itp"

[ moleculetype ]
; Name            nrexcl
et            3

'''
et_file.write(head)

Тип атома был взят из таблицы, полученной следующей командой в терминале:

cat /usr/share/gromacs/top/oplsaa.ff/atomtypes.atp

Я выбрала вполне подходящие типы для атомов C (opls_139) и H (opls_140).

## атомы
atoms = '''[ atoms ]
;   nr  type  resnr  residue  atom   cgnr     charge       mass
    1   opls_139      1    ETH      C1      1    -0.189      12.01
    2   opls_139      1    ETH      C2      2    -0.155      12.01
    3   opls_140      1    ETH      H1      3     0.0059       1.008
    4   opls_140      1    ETH      H2      4     0.0059       1.008
    5   opls_140      1    ETH      H3      5     0.0059       1.008
    6   opls_140      1    ETH      H4      6     0.0056       1.008
    7   opls_140      1    ETH      H5      7     0.0056       1.008
    8   opls_140      1    ETH      H6      8     0.0056       1.008
    
'''
et_file.write(atoms)

## связи
h = '''[ bonds ]
;  ai    aj funct  b0       kb
'''
et_file.write(h)
bonds = m3d.GetBonds()
for i,b in enumerate(bonds):
    l = '    ' + str(b.GetBeginAtomIdx() + 1) + '     ' + str(b.GetEndAtomIdx() + 1) + '     ' + str(1) + '\n'
    et_file.write(l)

## углы
h = '''[ angles ]
;  ai    aj    ak funct  phi0   kphi
'''
et_file.write(h) 
for b1 in m3d.GetBonds():
    for b2 in m3d.GetBonds():
        if b1.GetBeginAtomIdx() == b2.GetBeginAtomIdx() and b1.GetEndAtomIdx() < b2.GetEndAtomIdx():
            l = '    ' + str(b1.GetEndAtomIdx() + 1) + '     ' + str(b1.GetBeginAtomIdx() + 1) + '     ' +str(b2.GetEndAtomIdx() + 1)+'     '+ str(1) + '\n'
            et_file.write(l)

## торсионные углы
h = '''[ dihedrals ]
;  ai    aj    ak    al funct
'''
et_file.write(h) 
for b1 in m3d.GetBonds():
    for b2 in m3d.GetBonds():
        for b3 in m3d.GetBonds():
            if b1.GetBeginAtomIdx() == b2.GetBeginAtomIdx() and b1.GetEndAtomIdx() < b2.GetEndAtomIdx():
                if b1.GetEndAtomIdx() == b3.GetBeginAtomIdx():
                    l = '    '+str(b3.GetEndAtomIdx()+1)+'     ' + str(b1.GetEndAtomIdx() + 1) + '     ' + str(b1.GetBeginAtomIdx() + 1) + '     ' +str(b2.GetEndAtomIdx() + 1)+'     '+ str(3) + '\n'
                    et_file.write(l)

## пары
h = '''[ pairs ]
; список атомов 1-4
;  ai    aj funct
   3  6
   3  7
   3  8
   4  6
   4  7
   4  8
   5  6
   5  7
   5  8

[ System ]
; any text here
first one
[ molecules ]
;Name count
 et    1'''
et_file.write(h)
et_file.close()

У нас есть 5 файлов с различными параметрами температуры. Интересно посмотреть, какой лучше работает для нашей структуры. Запускаем для каждого gromp и mdrun.

import subprocess
for name in ['be','vr','nh','an','sd']:
    cmd = 'grompp -f %s.mdp -c etane.gro -p et.top -o et_%s.tpr >& log.grompp' %(name, name)
    subprocess.call(cmd, shell=True)

for filename in ['be','vr','nh','an','sd']:
    cmd = 'mdrun -deffnm et_%s.tpr -v -nt 1' %filename
    subprocess.call(cmd,shell=True)

Затем посмотрим наглядно на результаты. Для этого конвертируем результаты в pdb.

for filename in ['be','vr','nh','an','sd']:
    cmd = 'echo 0 | trjconv -f et_%s.tpr.trr -s et_%s.tpr -o et_%s.pdb >& log.trjconv' %(filename,filename,filename)
    subprocess.call(cmd,shell=True)

Тепловые колебания et_an почти отсутствуют, а у et_sd самые беспорядочные с неожиданными выбросами далеко от изначального центра молекулы. et_be, et_nh и et_vr крутятся вокруг C-C связи и еще дополнительно в разных плоскостях с разной амплитудой.

Теперь сравним результаты более научным способом. Для этого сравним потенциальную энергию связи и кинетическую энергию для каждой из 5 систем.

for filename in ['be','vr','nh','an','sd']:
    cmd1 = 'echo 10 | g_energy -f et_%s.tpr.edr -o et_%s_pot_en.xvg -xvg none >& log.g_energy' %(filename,filename)
    cmd2 = 'echo 11 | g_energy -f et_%s.tpr.edr -o et_%s_kin_en.xvg -xvg none >& log.g_energy' %(filename,filename)
    subprocess.call(cmd1,shell=True)
    subprocess.call(cmd2,shell=True)

Построим графики

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
for filename in ['be','vr','nh','an','sd']:
    a= np.loadtxt('et_%s_pot_en.xvg' % filename)
    b= np.loadtxt('et_%s_kin_en.xvg' % filename)
    t1=a[:,0]
    y1=a[:,1]
    t2=b[:,0]
    y2=b[:,1]
    plt.plot(t1, y1, "b-", t2, y2, "r-" )
    red_patch = mpatches.Patch(color='red', label='Kinetic')
    blue_patch = mpatches.Patch(color='blue', label='Potential')
    plt.legend(handles=[red_patch,blue_patch])
    plt.title('Energies of %s system' % filename) 
    name = filename + '.png'
    plt.savefig(name)

Интересно, что это сравнение выделяет систему et_be, так как только у нее колебания стремительно затухают. Амплитуда сильно разнится (наименьшая у et_an, наибольшая у et_nh, разница на 2 порядка).

Теперь посмотрим на колебание длины C-C связи. Для этого создадим индекс файл с одной связью b.ndx с индексами атомов, образующих связь (1 и 2 для этана). А затем построим распределения длин.

import subprocess
for filename in ['be','vr','nh','an','sd']:
    cmd = 'g_bond -f et_%s.tpr.trr -s et_%s.tpr -o bond_%s.xvg -n b.ndx -xvg none' % (filename,filename,filename)
    subprocess.call(cmd,shell=True)

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
for filename in ['be','vr','nh','an','sd']:
    a= np.loadtxt('bond_%s.xvg' % filename)
    t1=a[:,0]
    y1=a[:,1]
    width = 0.0001 #эмпирически
    plt.bar(t1,y1,width, color="blue")
    plt.title('Length of C-C bond in %s system' % filename)
    name = filename + 'CC.png'
    plt.savefig(name)

По форме на распределение Больцмана похожи только be и nh. максимум попадает на 1.53, что соответствует действительности. Остальные распределения не только не похожи на распределение Больцмана, но и имеют отличающиеся от 1.53 максимумы, а значению 1.53 неправдоподобно мало (как, например, у an). nh лучше всего подходит, как как не симметрично полностью, а имеет правый хвост длиннее левого.

Теперь сравним время работы алгоритма. Эту инфомацию можно получить из файла выдачи GROMACS, поле Time:

import subprocess
for filename in ['be','vr','nh','an','sd']:
    cmd1 = 'grompp -f %s.mdp -c etane.gro -p et.top -o et_%s.tpr >& log.grompp' %(filename,filename)
    cmd2 = 'mdrun -deffnm et_%s.tpr -v -nt 1' %filename
    output_file = open(filename+'.sh','w')
    output_file.writelines('\n'.join([cmd1,cmd2]))
    output_file.close()

%%bash
chmod +x *.sh
./be.sh
./vr.sh
./nh.sh
./an.sh
./sd.sh

                         :-)  G  R  O  M  A  C  S  (-:

                              S  C  A  M  O  R  G

                            :-)  VERSION 4.5.5  (-:

        Written by Emile Apol, Rossen Apostolov, Herman J.C. Berendsen,
      Aldert van Buuren, Pär Bjelkmar, Rudi van Drunen, Anton Feenstra, 
        Gerrit Groenhof, Peter Kasson, Per Larsson, Pieter Meulenhoff, 
           Teemu Murtola, Szilard Pall, Sander Pronk, Roland Schulz, 
                Michael Shirts, Alfons Sijbers, Peter Tieleman,

               Berk Hess, David van der Spoel, and Erik Lindahl.

       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
            Copyright (c) 2001-2010, The GROMACS development team at
        Uppsala University & The Royal Institute of Technology, Sweden.
            check out http://www.gromacs.org for more information.

         This program is free software; you can redistribute it and/or
          modify it under the terms of the GNU General Public License
         as published by the Free Software Foundation; either version 2
             of the License, or (at your option) any later version.

                                :-)  mdrun  (-:

Option     Filename  Type         Description
------------------------------------------------------------
  -s      et_be.tpr  Input        Run input file: tpr tpb tpa
  -o  et_be.tpr.trr  Output       Full precision trajectory: trr trj cpt
  -x  et_be.tpr.xtc  Output, Opt. Compressed trajectory (portable xdr format)
-cpi  et_be.tpr.cpt  Input, Opt.  Checkpoint file
-cpo  et_be.tpr.cpt  Output, Opt. Checkpoint file
  -c  et_be.tpr.gro  Output       Structure file: gro g96 pdb etc.
  -e  et_be.tpr.edr  Output       Energy file
  -g  et_be.tpr.log  Output       Log file
-dhdl et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
-field  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
-table  et_be.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tablep  et_be.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tableb  et_be.tpr.xvg  Input, Opt.  xvgr/xmgr file
-rerun  et_be.tpr.trr  Input, Opt.  Trajectory: xtc trr trj gro g96 pdb cpt
-tpi  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
-tpid et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
 -ei  et_be.tpr.edi  Input, Opt.  ED sampling input
 -eo  et_be.tpr.edo  Output, Opt. ED sampling output
  -j  et_be.tpr.gct  Input, Opt.  General coupling stuff
 -jo  et_be.tpr.gct  Output, Opt. General coupling stuff
-ffout  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
-devout  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
-runav  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
 -px  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
 -pf  et_be.tpr.xvg  Output, Opt. xvgr/xmgr file
-mtx  et_be.tpr.mtx  Output, Opt. Hessian matrix
 -dn  et_be.tpr.ndx  Output, Opt. Index file
-multidir et_be.tpr  Input, Opt., Mult. Run directory

Option       Type   Value   Description
------------------------------------------------------
-[no]h       bool   no      Print help info and quit
-[no]version bool   no      Print version info and quit
-nice        int    0       Set the nicelevel
-deffnm      string et_be.tpr  Set the default filename for all file options
-xvg         enum   xmgrace  xvg plot formatting: xmgrace, xmgr or none
-[no]pd      bool   no      Use particle decompostion
-dd          vector 0 0 0   Domain decomposition grid, 0 is optimize
-nt          int    1       Number of threads to start (0 is guess)
-npme        int    -1      Number of separate nodes to be used for PME, -1
                            is guess
-ddorder     enum   interleave  DD node order: interleave, pp_pme or cartesian
-[no]ddcheck bool   yes     Check for all bonded interactions with DD
-rdd         real   0       The maximum distance for bonded interactions with
                            DD (nm), 0 is determine from initial coordinates
-rcon        real   0       Maximum distance for P-LINCS (nm), 0 is estimate
-dlb         enum   auto    Dynamic load balancing (with DD): auto, no or yes
-dds         real   0.8     Minimum allowed dlb scaling of the DD cell size
-gcom        int    -1      Global communication frequency
-[no]v       bool   yes     Be loud and noisy
-[no]compact bool   yes     Write a compact log file
-[no]seppot  bool   no      Write separate V and dVdl terms for each
                            interaction type and node to the log file(s)
-pforce      real   -1      Print all forces larger than this (kJ/mol nm)
-[no]reprod  bool   no      Try to avoid optimizations that affect binary
                            reproducibility
-cpt         real   15      Checkpoint interval (minutes)
-[no]cpnum   bool   no      Keep and number checkpoint files
-[no]append  bool   yes     Append to previous output files when continuing
                            from checkpoint instead of adding the simulation
                            part number to all file names
-maxh        real   -1      Terminate after 0.99 times this time (hours)
-multi       int    0       Do multiple simulations in parallel
-replex      int    0       Attempt replica exchange periodically with this
                            period (steps)
-reseed      int    -1      Seed for replica exchange, -1 is generate a seed
-[no]ionize  bool   no      Do a simulation including the effect of an X-Ray
                            bombardment on your system


Back Off! I just backed up et_be.tpr.log to ./#et_be.tpr.log.2#
Getting Loaded...
Reading file et_be.tpr, VERSION 4.5.5 (single precision)
Loaded with Money


Back Off! I just backed up et_be.tpr.trr to ./#et_be.tpr.trr.2#

Back Off! I just backed up et_be.tpr.edr to ./#et_be.tpr.edr.2#
starting mdrun 'first one'
250000 steps,    250.0 ps.
step 249900, remaining runtime:     0 s          
Writing final coordinates.

Back Off! I just backed up et_be.tpr.gro to ./#et_be.tpr.gro.2#
step 250000, remaining runtime:     0 s          
               NODE (s)   Real (s)      (%)
       Time:      3.890      4.147     93.8
               (Mnbf/s)   (MFlops)   (ns/day)  (hour/ns)
Performance:      0.000    350.890   5552.722      0.004

gcq#15: "Don't Grumble, Give a Whistle !" (Monty Python)

                         :-)  G  R  O  M  A  C  S  (-:

               GRoups of Organic Molecules in ACtion for Science

                            :-)  VERSION 4.5.5  (-:

        Written by Emile Apol, Rossen Apostolov, Herman J.C. Berendsen,
      Aldert van Buuren, Pär Bjelkmar, Rudi van Drunen, Anton Feenstra, 
        Gerrit Groenhof, Peter Kasson, Per Larsson, Pieter Meulenhoff, 
           Teemu Murtola, Szilard Pall, Sander Pronk, Roland Schulz, 
                Michael Shirts, Alfons Sijbers, Peter Tieleman,

               Berk Hess, David van der Spoel, and Erik Lindahl.

       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
            Copyright (c) 2001-2010, The GROMACS development team at
        Uppsala University & The Royal Institute of Technology, Sweden.
            check out http://www.gromacs.org for more information.

         This program is free software; you can redistribute it and/or
          modify it under the terms of the GNU General Public License
         as published by the Free Software Foundation; either version 2
             of the License, or (at your option) any later version.

                                :-)  mdrun  (-:

Option     Filename  Type         Description
------------------------------------------------------------
  -s      et_vr.tpr  Input        Run input file: tpr tpb tpa
  -o  et_vr.tpr.trr  Output       Full precision trajectory: trr trj cpt
  -x  et_vr.tpr.xtc  Output, Opt. Compressed trajectory (portable xdr format)
-cpi  et_vr.tpr.cpt  Input, Opt.  Checkpoint file
-cpo  et_vr.tpr.cpt  Output, Opt. Checkpoint file
  -c  et_vr.tpr.gro  Output       Structure file: gro g96 pdb etc.
  -e  et_vr.tpr.edr  Output       Energy file
  -g  et_vr.tpr.log  Output       Log file
-dhdl et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
-field  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
-table  et_vr.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tablep  et_vr.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tableb  et_vr.tpr.xvg  Input, Opt.  xvgr/xmgr file
-rerun  et_vr.tpr.trr  Input, Opt.  Trajectory: xtc trr trj gro g96 pdb cpt
-tpi  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
-tpid et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
 -ei  et_vr.tpr.edi  Input, Opt.  ED sampling input
 -eo  et_vr.tpr.edo  Output, Opt. ED sampling output
  -j  et_vr.tpr.gct  Input, Opt.  General coupling stuff
 -jo  et_vr.tpr.gct  Output, Opt. General coupling stuff
-ffout  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
-devout  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
-runav  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
 -px  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
 -pf  et_vr.tpr.xvg  Output, Opt. xvgr/xmgr file
-mtx  et_vr.tpr.mtx  Output, Opt. Hessian matrix
 -dn  et_vr.tpr.ndx  Output, Opt. Index file
-multidir et_vr.tpr  Input, Opt., Mult. Run directory

Option       Type   Value   Description
------------------------------------------------------
-[no]h       bool   no      Print help info and quit
-[no]version bool   no      Print version info and quit
-nice        int    0       Set the nicelevel
-deffnm      string et_vr.tpr  Set the default filename for all file options
-xvg         enum   xmgrace  xvg plot formatting: xmgrace, xmgr or none
-[no]pd      bool   no      Use particle decompostion
-dd          vector 0 0 0   Domain decomposition grid, 0 is optimize
-nt          int    1       Number of threads to start (0 is guess)
-npme        int    -1      Number of separate nodes to be used for PME, -1
                            is guess
-ddorder     enum   interleave  DD node order: interleave, pp_pme or cartesian
-[no]ddcheck bool   yes     Check for all bonded interactions with DD
-rdd         real   0       The maximum distance for bonded interactions with
                            DD (nm), 0 is determine from initial coordinates
-rcon        real   0       Maximum distance for P-LINCS (nm), 0 is estimate
-dlb         enum   auto    Dynamic load balancing (with DD): auto, no or yes
-dds         real   0.8     Minimum allowed dlb scaling of the DD cell size
-gcom        int    -1      Global communication frequency
-[no]v       bool   yes     Be loud and noisy
-[no]compact bool   yes     Write a compact log file
-[no]seppot  bool   no      Write separate V and dVdl terms for each
                            interaction type and node to the log file(s)
-pforce      real   -1      Print all forces larger than this (kJ/mol nm)
-[no]reprod  bool   no      Try to avoid optimizations that affect binary
                            reproducibility
-cpt         real   15      Checkpoint interval (minutes)
-[no]cpnum   bool   no      Keep and number checkpoint files
-[no]append  bool   yes     Append to previous output files when continuing
                            from checkpoint instead of adding the simulation
                            part number to all file names
-maxh        real   -1      Terminate after 0.99 times this time (hours)
-multi       int    0       Do multiple simulations in parallel
-replex      int    0       Attempt replica exchange periodically with this
                            period (steps)
-reseed      int    -1      Seed for replica exchange, -1 is generate a seed
-[no]ionize  bool   no      Do a simulation including the effect of an X-Ray
                            bombardment on your system


Back Off! I just backed up et_vr.tpr.log to ./#et_vr.tpr.log.2#
Getting Loaded...
Reading file et_vr.tpr, VERSION 4.5.5 (single precision)
Loaded with Money


Back Off! I just backed up et_vr.tpr.trr to ./#et_vr.tpr.trr.2#

Back Off! I just backed up et_vr.tpr.edr to ./#et_vr.tpr.edr.2#
starting mdrun 'first one'
250000 steps,    250.0 ps.
step 249900, remaining runtime:     0 s          
Writing final coordinates.

Back Off! I just backed up et_vr.tpr.gro to ./#et_vr.tpr.gro.2#
step 250000, remaining runtime:     0 s          
               NODE (s)   Real (s)      (%)
       Time:      3.840      4.006     95.9
               (Mnbf/s)   (MFlops)   (ns/day)  (hour/ns)
Performance:      0.000    355.393   5625.023      0.004

gcq#109: "Ease Myself Into the Body Bag" (P.J. Harvey)

                         :-)  G  R  O  M  A  C  S  (-:

                          GROtesk MACabre and Sinister

                            :-)  VERSION 4.5.5  (-:

        Written by Emile Apol, Rossen Apostolov, Herman J.C. Berendsen,
      Aldert van Buuren, Pär Bjelkmar, Rudi van Drunen, Anton Feenstra, 
        Gerrit Groenhof, Peter Kasson, Per Larsson, Pieter Meulenhoff, 
           Teemu Murtola, Szilard Pall, Sander Pronk, Roland Schulz, 
                Michael Shirts, Alfons Sijbers, Peter Tieleman,

               Berk Hess, David van der Spoel, and Erik Lindahl.

       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
            Copyright (c) 2001-2010, The GROMACS development team at
        Uppsala University & The Royal Institute of Technology, Sweden.
            check out http://www.gromacs.org for more information.

         This program is free software; you can redistribute it and/or
          modify it under the terms of the GNU General Public License
         as published by the Free Software Foundation; either version 2
             of the License, or (at your option) any later version.

                                :-)  mdrun  (-:

Option     Filename  Type         Description
------------------------------------------------------------
  -s      et_nh.tpr  Input        Run input file: tpr tpb tpa
  -o  et_nh.tpr.trr  Output       Full precision trajectory: trr trj cpt
  -x  et_nh.tpr.xtc  Output, Opt. Compressed trajectory (portable xdr format)
-cpi  et_nh.tpr.cpt  Input, Opt.  Checkpoint file
-cpo  et_nh.tpr.cpt  Output, Opt. Checkpoint file
  -c  et_nh.tpr.gro  Output       Structure file: gro g96 pdb etc.
  -e  et_nh.tpr.edr  Output       Energy file
  -g  et_nh.tpr.log  Output       Log file
-dhdl et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
-field  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
-table  et_nh.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tablep  et_nh.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tableb  et_nh.tpr.xvg  Input, Opt.  xvgr/xmgr file
-rerun  et_nh.tpr.trr  Input, Opt.  Trajectory: xtc trr trj gro g96 pdb cpt
-tpi  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
-tpid et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
 -ei  et_nh.tpr.edi  Input, Opt.  ED sampling input
 -eo  et_nh.tpr.edo  Output, Opt. ED sampling output
  -j  et_nh.tpr.gct  Input, Opt.  General coupling stuff
 -jo  et_nh.tpr.gct  Output, Opt. General coupling stuff
-ffout  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
-devout  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
-runav  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
 -px  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
 -pf  et_nh.tpr.xvg  Output, Opt. xvgr/xmgr file
-mtx  et_nh.tpr.mtx  Output, Opt. Hessian matrix
 -dn  et_nh.tpr.ndx  Output, Opt. Index file
-multidir et_nh.tpr  Input, Opt., Mult. Run directory

Option       Type   Value   Description
------------------------------------------------------
-[no]h       bool   no      Print help info and quit
-[no]version bool   no      Print version info and quit
-nice        int    0       Set the nicelevel
-deffnm      string et_nh.tpr  Set the default filename for all file options
-xvg         enum   xmgrace  xvg plot formatting: xmgrace, xmgr or none
-[no]pd      bool   no      Use particle decompostion
-dd          vector 0 0 0   Domain decomposition grid, 0 is optimize
-nt          int    1       Number of threads to start (0 is guess)
-npme        int    -1      Number of separate nodes to be used for PME, -1
                            is guess
-ddorder     enum   interleave  DD node order: interleave, pp_pme or cartesian
-[no]ddcheck bool   yes     Check for all bonded interactions with DD
-rdd         real   0       The maximum distance for bonded interactions with
                            DD (nm), 0 is determine from initial coordinates
-rcon        real   0       Maximum distance for P-LINCS (nm), 0 is estimate
-dlb         enum   auto    Dynamic load balancing (with DD): auto, no or yes
-dds         real   0.8     Minimum allowed dlb scaling of the DD cell size
-gcom        int    -1      Global communication frequency
-[no]v       bool   yes     Be loud and noisy
-[no]compact bool   yes     Write a compact log file
-[no]seppot  bool   no      Write separate V and dVdl terms for each
                            interaction type and node to the log file(s)
-pforce      real   -1      Print all forces larger than this (kJ/mol nm)
-[no]reprod  bool   no      Try to avoid optimizations that affect binary
                            reproducibility
-cpt         real   15      Checkpoint interval (minutes)
-[no]cpnum   bool   no      Keep and number checkpoint files
-[no]append  bool   yes     Append to previous output files when continuing
                            from checkpoint instead of adding the simulation
                            part number to all file names
-maxh        real   -1      Terminate after 0.99 times this time (hours)
-multi       int    0       Do multiple simulations in parallel
-replex      int    0       Attempt replica exchange periodically with this
                            period (steps)
-reseed      int    -1      Seed for replica exchange, -1 is generate a seed
-[no]ionize  bool   no      Do a simulation including the effect of an X-Ray
                            bombardment on your system


Back Off! I just backed up et_nh.tpr.log to ./#et_nh.tpr.log.2#
Getting Loaded...
Reading file et_nh.tpr, VERSION 4.5.5 (single precision)
Loaded with Money


Back Off! I just backed up et_nh.tpr.trr to ./#et_nh.tpr.trr.2#

Back Off! I just backed up et_nh.tpr.edr to ./#et_nh.tpr.edr.2#
starting mdrun 'first one'
250000 steps,    250.0 ps.
step 249900, remaining runtime:     0 s          
Writing final coordinates.

Back Off! I just backed up et_nh.tpr.gro to ./#et_nh.tpr.gro.2#
step 250000, remaining runtime:     0 s          
               NODE (s)   Real (s)      (%)
       Time:      3.870      4.038     95.8
               (Mnbf/s)   (MFlops)   (ns/day)  (hour/ns)
Performance:      0.000    352.435   5581.418      0.004

gcq#203: "Hangout In the Suburbs If You've Got the Guts" (Urban Dance Squad)

                         :-)  G  R  O  M  A  C  S  (-:

           Glycine aRginine prOline Methionine Alanine Cystine Serine

                            :-)  VERSION 4.5.5  (-:

        Written by Emile Apol, Rossen Apostolov, Herman J.C. Berendsen,
      Aldert van Buuren, Pär Bjelkmar, Rudi van Drunen, Anton Feenstra, 
        Gerrit Groenhof, Peter Kasson, Per Larsson, Pieter Meulenhoff, 
           Teemu Murtola, Szilard Pall, Sander Pronk, Roland Schulz, 
                Michael Shirts, Alfons Sijbers, Peter Tieleman,

               Berk Hess, David van der Spoel, and Erik Lindahl.

       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
            Copyright (c) 2001-2010, The GROMACS development team at
        Uppsala University & The Royal Institute of Technology, Sweden.
            check out http://www.gromacs.org for more information.

         This program is free software; you can redistribute it and/or
          modify it under the terms of the GNU General Public License
         as published by the Free Software Foundation; either version 2
             of the License, or (at your option) any later version.

                                :-)  mdrun  (-:

Option     Filename  Type         Description
------------------------------------------------------------
  -s      et_an.tpr  Input        Run input file: tpr tpb tpa
  -o  et_an.tpr.trr  Output       Full precision trajectory: trr trj cpt
  -x  et_an.tpr.xtc  Output, Opt. Compressed trajectory (portable xdr format)
-cpi  et_an.tpr.cpt  Input, Opt.  Checkpoint file
-cpo  et_an.tpr.cpt  Output, Opt. Checkpoint file
  -c  et_an.tpr.gro  Output       Structure file: gro g96 pdb etc.
  -e  et_an.tpr.edr  Output       Energy file
  -g  et_an.tpr.log  Output       Log file
-dhdl et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
-field  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
-table  et_an.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tablep  et_an.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tableb  et_an.tpr.xvg  Input, Opt.  xvgr/xmgr file
-rerun  et_an.tpr.trr  Input, Opt.  Trajectory: xtc trr trj gro g96 pdb cpt
-tpi  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
-tpid et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
 -ei  et_an.tpr.edi  Input, Opt.  ED sampling input
 -eo  et_an.tpr.edo  Output, Opt. ED sampling output
  -j  et_an.tpr.gct  Input, Opt.  General coupling stuff
 -jo  et_an.tpr.gct  Output, Opt. General coupling stuff
-ffout  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
-devout  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
-runav  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
 -px  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
 -pf  et_an.tpr.xvg  Output, Opt. xvgr/xmgr file
-mtx  et_an.tpr.mtx  Output, Opt. Hessian matrix
 -dn  et_an.tpr.ndx  Output, Opt. Index file
-multidir et_an.tpr  Input, Opt., Mult. Run directory

Option       Type   Value   Description
------------------------------------------------------
-[no]h       bool   no      Print help info and quit
-[no]version bool   no      Print version info and quit
-nice        int    0       Set the nicelevel
-deffnm      string et_an.tpr  Set the default filename for all file options
-xvg         enum   xmgrace  xvg plot formatting: xmgrace, xmgr or none
-[no]pd      bool   no      Use particle decompostion
-dd          vector 0 0 0   Domain decomposition grid, 0 is optimize
-nt          int    1       Number of threads to start (0 is guess)
-npme        int    -1      Number of separate nodes to be used for PME, -1
                            is guess
-ddorder     enum   interleave  DD node order: interleave, pp_pme or cartesian
-[no]ddcheck bool   yes     Check for all bonded interactions with DD
-rdd         real   0       The maximum distance for bonded interactions with
                            DD (nm), 0 is determine from initial coordinates
-rcon        real   0       Maximum distance for P-LINCS (nm), 0 is estimate
-dlb         enum   auto    Dynamic load balancing (with DD): auto, no or yes
-dds         real   0.8     Minimum allowed dlb scaling of the DD cell size
-gcom        int    -1      Global communication frequency
-[no]v       bool   yes     Be loud and noisy
-[no]compact bool   yes     Write a compact log file
-[no]seppot  bool   no      Write separate V and dVdl terms for each
                            interaction type and node to the log file(s)
-pforce      real   -1      Print all forces larger than this (kJ/mol nm)
-[no]reprod  bool   no      Try to avoid optimizations that affect binary
                            reproducibility
-cpt         real   15      Checkpoint interval (minutes)
-[no]cpnum   bool   no      Keep and number checkpoint files
-[no]append  bool   yes     Append to previous output files when continuing
                            from checkpoint instead of adding the simulation
                            part number to all file names
-maxh        real   -1      Terminate after 0.99 times this time (hours)
-multi       int    0       Do multiple simulations in parallel
-replex      int    0       Attempt replica exchange periodically with this
                            period (steps)
-reseed      int    -1      Seed for replica exchange, -1 is generate a seed
-[no]ionize  bool   no      Do a simulation including the effect of an X-Ray
                            bombardment on your system


Back Off! I just backed up et_an.tpr.log to ./#et_an.tpr.log.2#
Getting Loaded...
Reading file et_an.tpr, VERSION 4.5.5 (single precision)
Loaded with Money


Back Off! I just backed up et_an.tpr.trr to ./#et_an.tpr.trr.2#

Back Off! I just backed up et_an.tpr.edr to ./#et_an.tpr.edr.2#
starting mdrun 'first one'
250000 steps,    250.0 ps.
step 249900, remaining runtime:     0 s          
Writing final coordinates.

Back Off! I just backed up et_an.tpr.gro to ./#et_an.tpr.gro.2#
step 250000, remaining runtime:     0 s          
               NODE (s)   Real (s)      (%)
       Time:      3.760      3.974     94.6
               (Mnbf/s)   (MFlops)   (ns/day)  (hour/ns)
Performance:      0.000    363.599   5744.704      0.004

gcq#46: "I Wonder, Should I Get Up..." (J. Lennon)

                         :-)  G  R  O  M  A  C  S  (-:

                     Gyas ROwers Mature At Cryogenic Speed

                            :-)  VERSION 4.5.5  (-:

        Written by Emile Apol, Rossen Apostolov, Herman J.C. Berendsen,
      Aldert van Buuren, Pär Bjelkmar, Rudi van Drunen, Anton Feenstra, 
        Gerrit Groenhof, Peter Kasson, Per Larsson, Pieter Meulenhoff, 
           Teemu Murtola, Szilard Pall, Sander Pronk, Roland Schulz, 
                Michael Shirts, Alfons Sijbers, Peter Tieleman,

               Berk Hess, David van der Spoel, and Erik Lindahl.

       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
            Copyright (c) 2001-2010, The GROMACS development team at
        Uppsala University & The Royal Institute of Technology, Sweden.
            check out http://www.gromacs.org for more information.

         This program is free software; you can redistribute it and/or
          modify it under the terms of the GNU General Public License
         as published by the Free Software Foundation; either version 2
             of the License, or (at your option) any later version.

                                :-)  mdrun  (-:

Option     Filename  Type         Description
------------------------------------------------------------
  -s      et_sd.tpr  Input        Run input file: tpr tpb tpa
  -o  et_sd.tpr.trr  Output       Full precision trajectory: trr trj cpt
  -x  et_sd.tpr.xtc  Output, Opt. Compressed trajectory (portable xdr format)
-cpi  et_sd.tpr.cpt  Input, Opt.  Checkpoint file
-cpo  et_sd.tpr.cpt  Output, Opt. Checkpoint file
  -c  et_sd.tpr.gro  Output       Structure file: gro g96 pdb etc.
  -e  et_sd.tpr.edr  Output       Energy file
  -g  et_sd.tpr.log  Output       Log file
-dhdl et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
-field  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
-table  et_sd.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tablep  et_sd.tpr.xvg  Input, Opt.  xvgr/xmgr file
-tableb  et_sd.tpr.xvg  Input, Opt.  xvgr/xmgr file
-rerun  et_sd.tpr.trr  Input, Opt.  Trajectory: xtc trr trj gro g96 pdb cpt
-tpi  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
-tpid et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
 -ei  et_sd.tpr.edi  Input, Opt.  ED sampling input
 -eo  et_sd.tpr.edo  Output, Opt. ED sampling output
  -j  et_sd.tpr.gct  Input, Opt.  General coupling stuff
 -jo  et_sd.tpr.gct  Output, Opt. General coupling stuff
-ffout  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
-devout  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
-runav  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
 -px  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
 -pf  et_sd.tpr.xvg  Output, Opt. xvgr/xmgr file
-mtx  et_sd.tpr.mtx  Output, Opt. Hessian matrix
 -dn  et_sd.tpr.ndx  Output, Opt. Index file
-multidir et_sd.tpr  Input, Opt., Mult. Run directory

Option       Type   Value   Description
------------------------------------------------------
-[no]h       bool   no      Print help info and quit
-[no]version bool   no      Print version info and quit
-nice        int    0       Set the nicelevel
-deffnm      string et_sd.tpr  Set the default filename for all file options
-xvg         enum   xmgrace  xvg plot formatting: xmgrace, xmgr or none
-[no]pd      bool   no      Use particle decompostion
-dd          vector 0 0 0   Domain decomposition grid, 0 is optimize
-nt          int    1       Number of threads to start (0 is guess)
-npme        int    -1      Number of separate nodes to be used for PME, -1
                            is guess
-ddorder     enum   interleave  DD node order: interleave, pp_pme or cartesian
-[no]ddcheck bool   yes     Check for all bonded interactions with DD
-rdd         real   0       The maximum distance for bonded interactions with
                            DD (nm), 0 is determine from initial coordinates
-rcon        real   0       Maximum distance for P-LINCS (nm), 0 is estimate
-dlb         enum   auto    Dynamic load balancing (with DD): auto, no or yes
-dds         real   0.8     Minimum allowed dlb scaling of the DD cell size
-gcom        int    -1      Global communication frequency
-[no]v       bool   yes     Be loud and noisy
-[no]compact bool   yes     Write a compact log file
-[no]seppot  bool   no      Write separate V and dVdl terms for each
                            interaction type and node to the log file(s)
-pforce      real   -1      Print all forces larger than this (kJ/mol nm)
-[no]reprod  bool   no      Try to avoid optimizations that affect binary
                            reproducibility
-cpt         real   15      Checkpoint interval (minutes)
-[no]cpnum   bool   no      Keep and number checkpoint files
-[no]append  bool   yes     Append to previous output files when continuing
                            from checkpoint instead of adding the simulation
                            part number to all file names
-maxh        real   -1      Terminate after 0.99 times this time (hours)
-multi       int    0       Do multiple simulations in parallel
-replex      int    0       Attempt replica exchange periodically with this
                            period (steps)
-reseed      int    -1      Seed for replica exchange, -1 is generate a seed
-[no]ionize  bool   no      Do a simulation including the effect of an X-Ray
                            bombardment on your system


Back Off! I just backed up et_sd.tpr.log to ./#et_sd.tpr.log.2#
Getting Loaded...
Reading file et_sd.tpr, VERSION 4.5.5 (single precision)
Loaded with Money


Back Off! I just backed up et_sd.tpr.trr to ./#et_sd.tpr.trr.2#

Back Off! I just backed up et_sd.tpr.edr to ./#et_sd.tpr.edr.2#
starting mdrun 'first one'
250000 steps,    250.0 ps.
step 249900, remaining runtime:     0 s          
Writing final coordinates.

Back Off! I just backed up et_sd.tpr.gro to ./#et_sd.tpr.gro.2#
step 250000, remaining runtime:     0 s          
               NODE (s)   Real (s)      (%)
       Time:      4.470      4.673     95.7
               (Mnbf/s)   (MFlops)   (ns/day)  (hour/ns)
Performance:      0.000    319.124   4832.234      0.005

gcq#255: "This is Tense !" (Star Wars Episode I The Phantom Menace)

Видно, что an оказался быстрее всех, а sd дольше всех. это коррелирует с отклонениями в координатах (et_an совсем почти не колебалась, а et_sd -- очень сильно).

Таблица 1. Время работы алгоритма.

Algorithm	Time, s
an	3.730
be	3.780
nh	3.890
sd	4.400
vr	3.860