2 See the "run control" section for a working example of the
3 syntax to use when making .mdp entries, with and without detailed
4 documentation for values those entries might take. Everything can
5 be cross-referenced, see the examples there. TODO Make more
8 Molecular dynamics parameters (.mdp options)
9 ============================================
16 Default values are given in parentheses, or listed first among
17 choices. The first option in the list is always the default
18 option. Units are given in square brackets. The difference between a
19 dash and an underscore is ignored.
21 A :ref:`sample mdp file <mdp>` is available. This should be
22 appropriate to start a normal simulation. Edit it to suit your
23 specific needs and desires.
31 directories to include in your topology. Format:
32 ``-I/home/john/mylib -I../otherlib``
36 defines to pass to the preprocessor, default is no defines. You can
37 use any defines to control options in your customized topology
38 files. Options that act on existing :ref:`top` file mechanisms
41 ``-DFLEXIBLE`` will use flexible water instead of rigid water
42 into your topology, this can be useful for normal mode analysis.
44 ``-DPOSRES`` will trigger the inclusion of ``posre.itp`` into
45 your topology, used for implementing position restraints.
53 (Despite the name, this list includes algorithms that are not
54 actually integrators over time. :mdp-value:`integrator=steep` and
55 all entries following it are in this category)
59 A leap-frog algorithm for integrating Newton's equations of motion.
63 A velocity Verlet algorithm for integrating Newton's equations
64 of motion. For constant NVE simulations started from
65 corresponding points in the same trajectory, the trajectories
66 are analytically, but not binary, identical to the
67 :mdp-value:`integrator=md` leap-frog integrator. The the kinetic
68 energy, which is determined from the whole step velocities and
69 is therefore slightly too high. The advantage of this integrator
70 is more accurate, reversible Nose-Hoover and Parrinello-Rahman
71 coupling integration based on Trotter expansion, as well as
72 (slightly too small) full step velocity output. This all comes
73 at the cost off extra computation, especially with constraints
74 and extra communication in parallel. Note that for nearly all
75 production simulations the :mdp-value:`integrator=md` integrator
78 .. mdp-value:: md-vv-avek
80 A velocity Verlet algorithm identical to
81 :mdp-value:`integrator=md-vv`, except that the kinetic energy is
82 determined as the average of the two half step kinetic energies
83 as in the :mdp-value:`integrator=md` integrator, and this thus
84 more accurate. With Nose-Hoover and/or Parrinello-Rahman
85 coupling this comes with a slight increase in computational
90 An accurate and efficient leap-frog stochastic dynamics
91 integrator. With constraints, coordinates needs to be
92 constrained twice per integration step. Depending on the
93 computational cost of the force calculation, this can take a
94 significant part of the simulation time. The temperature for one
95 or more groups of atoms (:mdp:`tc-grps`) is set with
96 :mdp:`ref-t`, the inverse friction constant for each group is
97 set with :mdp:`tau-t`. The parameter :mdp:`tcoupl` is
98 ignored. The random generator is initialized with
99 :mdp:`ld-seed`. When used as a thermostat, an appropriate value
100 for :mdp:`tau-t` is 2 ps, since this results in a friction that
101 is lower than the internal friction of water, while it is high
102 enough to remove excess heat NOTE: temperature deviations decay
103 twice as fast as with a Berendsen thermostat with the same
108 An Euler integrator for Brownian or position Langevin dynamics,
109 the velocity is the force divided by a friction coefficient
110 (:mdp:`bd-fric`) plus random thermal noise (:mdp:`ref-t`). When
111 :mdp:`bd-fric` is 0, the friction coefficient for each particle
112 is calculated as mass/ :mdp:`tau-t`, as for the integrator
113 :mdp-value:`integrator=sd`. The random generator is initialized
118 A steepest descent algorithm for energy minimization. The
119 maximum step size is :mdp:`emstep`, the tolerance is
124 A conjugate gradient algorithm for energy minimization, the
125 tolerance is :mdp:`emtol`. CG is more efficient when a steepest
126 descent step is done every once in a while, this is determined
127 by :mdp:`nstcgsteep`. For a minimization prior to a normal mode
128 analysis, which requires a very high accuracy, |Gromacs| should be
129 compiled in double precision.
131 .. mdp-value:: l-bfgs
133 A quasi-Newtonian algorithm for energy minimization according to
134 the low-memory Broyden-Fletcher-Goldfarb-Shanno approach. In
135 practice this seems to converge faster than Conjugate Gradients,
136 but due to the correction steps necessary it is not (yet)
141 Normal mode analysis is performed on the structure in the :ref:`tpr`
142 file. |Gromacs| should be compiled in double precision.
146 Test particle insertion. The last molecule in the topology is
147 the test particle. A trajectory must be provided to ``mdrun
148 -rerun``. This trajectory should not contain the molecule to be
149 inserted. Insertions are performed :mdp:`nsteps` times in each
150 frame at random locations and with random orientiations of the
151 molecule. When :mdp:`nstlist` is larger than one,
152 :mdp:`nstlist` insertions are performed in a sphere with radius
153 :mdp:`rtpi` around a the same random location using the same
154 pair list. Since pair list construction is expensive,
155 one can perform several extra insertions with the same list
156 almost for free. The random seed is set with
157 :mdp:`ld-seed`. The temperature for the Boltzmann weighting is
158 set with :mdp:`ref-t`, this should match the temperature of the
159 simulation of the original trajectory. Dispersion correction is
160 implemented correctly for TPI. All relevant quantities are
161 written to the file specified with ``mdrun -tpi``. The
162 distribution of insertion energies is written to the file
163 specified with ``mdrun -tpid``. No trajectory or energy file is
164 written. Parallel TPI gives identical results to single-node
165 TPI. For charged molecules, using PME with a fine grid is most
166 accurate and also efficient, since the potential in the system
167 only needs to be calculated once per frame.
171 Test particle insertion into a predefined cavity location. The
172 procedure is the same as for :mdp-value:`integrator=tpi`, except
173 that one coordinate extra is read from the trajectory, which is
174 used as the insertion location. The molecule to be inserted
175 should be centered at 0,0,0. |Gromacs| does not do this for you,
176 since for different situations a different way of centering
177 might be optimal. Also :mdp:`rtpi` sets the radius for the
178 sphere around this location. Neighbor searching is done only
179 once per frame, :mdp:`nstlist` is not used. Parallel
180 :mdp-value:`integrator=tpic` gives identical results to
181 single-rank :mdp-value:`integrator=tpic`.
185 Enable MiMiC QM/MM coupling to run hybrid molecular dynamics.
186 Keey in mind that its required to launch CPMD compiled with MiMiC as well.
187 In this mode all options regarding integration (T-coupling, P-coupling,
188 timestep and number of steps) are ignored as CPMD will do the integration
189 instead. Options related to forces computation (cutoffs, PME parameters,
190 etc.) are working as usual. Atom selection to define QM atoms is read
191 from :mdp:`QMMM-grps`
196 starting time for your run (only makes sense for time-based
202 time step for integration (only makes sense for time-based
208 maximum number of steps to integrate or minimize, -1 is no
214 The starting step. The time at step i in a run is
215 calculated as: t = :mdp:`tinit` + :mdp:`dt` *
216 (:mdp:`init-step` + i). The free-energy lambda is calculated
217 as: lambda = :mdp:`init-lambda` + :mdp:`delta-lambda` *
218 (:mdp:`init-step` + i). Also non-equilibrium MD parameters can
219 depend on the step number. Thus for exact restarts or redoing
220 part of a run it might be necessary to set :mdp:`init-step` to
221 the step number of the restart frame. :ref:`gmx convert-tpr`
222 does this automatically.
224 .. mdp:: simulation-part
227 A simulation can consist of multiple parts, each of which has
228 a part number. This option specifies what that number will
229 be, which helps keep track of parts that are logically the
230 same simulation. This option is generally useful to set only
231 when coping with a crashed simulation where files were lost.
235 .. mdp-value:: Linear
237 Remove center of mass translational velocity
239 .. mdp-value:: Angular
241 Remove center of mass translational and rotational velocity
243 .. mdp-value:: Linear-acceleration-correction
245 Remove center of mass translational velocity. Correct the center of
246 mass position assuming linear acceleration over :mdp:`nstcomm` steps.
247 This is useful for cases where an acceleration is expected on the
248 center of mass which is nearly constant over :mdp:`nstcomm` steps.
249 This can occur for example when pulling on a group using an absolute
254 No restriction on the center of mass motion
259 frequency for center of mass motion removal
263 group(s) for center of mass motion removal, default is the whole
272 (0) [amu ps\ :sup:`-1`]
273 Brownian dynamics friction coefficient. When :mdp:`bd-fric` is 0,
274 the friction coefficient for each particle is calculated as mass/
280 used to initialize random generator for thermal noise for
281 stochastic and Brownian dynamics. When :mdp:`ld-seed` is set to -1,
282 a pseudo random seed is used. When running BD or SD on multiple
283 processors, each processor uses a seed equal to :mdp:`ld-seed` plus
284 the processor number.
292 (10.0) [kJ mol\ :sup:`-1` nm\ :sup:`-1`]
293 the minimization is converged when the maximum force is smaller
304 frequency of performing 1 steepest descent step while doing
305 conjugate gradient energy minimization.
310 Number of correction steps to use for L-BFGS minimization. A higher
311 number is (at least theoretically) more accurate, but slower.
314 Shell Molecular Dynamics
315 ^^^^^^^^^^^^^^^^^^^^^^^^
317 When shells or flexible constraints are present in the system the
318 positions of the shells and the lengths of the flexible constraints
319 are optimized at every time step until either the RMS force on the
320 shells and constraints is less than :mdp:`emtol`, or a maximum number
321 of iterations :mdp:`niter` has been reached. Minimization is converged
322 when the maximum force is smaller than :mdp:`emtol`. For shell MD this
323 value should be 1.0 at most.
328 maximum number of iterations for optimizing the shell positions and
329 the flexible constraints.
334 the step size for optimizing the flexible constraints. Should be
335 chosen as mu/(d2V/dq2) where mu is the reduced mass of two
336 particles in a flexible constraint and d2V/dq2 is the second
337 derivative of the potential in the constraint direction. Hopefully
338 this number does not differ too much between the flexible
339 constraints, as the number of iterations and thus the runtime is
340 very sensitive to fcstep. Try several values!
343 Test particle insertion
344 ^^^^^^^^^^^^^^^^^^^^^^^
349 the test particle insertion radius, see integrators
350 :mdp-value:`integrator=tpi` and :mdp-value:`integrator=tpic`
359 number of steps that elapse between writing coordinates to the output
360 trajectory file (:ref:`trr`), the last coordinates are always written
365 number of steps that elapse between writing velocities to the output
366 trajectory file (:ref:`trr`), the last velocities are always written
371 number of steps that elapse between writing forces to the output
372 trajectory file (:ref:`trr`), the last forces are always written.
377 number of steps that elapse between writing energies to the log
378 file, the last energies are always written
380 .. mdp:: nstcalcenergy
383 number of steps that elapse between calculating the energies, 0 is
384 never. This option is only relevant with dynamics. This option affects the
385 performance in parallel simulations, because calculating energies
386 requires global communication between all processes which can
387 become a bottleneck at high parallelization.
392 number of steps that elapse between writing energies to energy file,
393 the last energies are always written, should be a multiple of
394 :mdp:`nstcalcenergy`. Note that the exact sums and fluctuations
395 over all MD steps modulo :mdp:`nstcalcenergy` are stored in the
396 energy file, so :ref:`gmx energy` can report exact energy averages
397 and fluctuations also when :mdp:`nstenergy` > 1
399 .. mdp:: nstxout-compressed
402 number of steps that elapse between writing position coordinates
403 using lossy compression (:ref:`xtc` file)
405 .. mdp:: compressed-x-precision
408 precision with which to write to the compressed trajectory file
410 .. mdp:: compressed-x-grps
412 group(s) to write to the compressed trajectory file, by default the
413 whole system is written (if :mdp:`nstxout-compressed` > 0)
417 group(s) for which to write to write short-ranged non-bonded
418 potential energies to the energy file (not supported on GPUs)
424 .. mdp:: cutoff-scheme
426 .. mdp-value:: Verlet
428 Generate a pair list with buffering. The buffer size is
429 automatically set based on :mdp:`verlet-buffer-tolerance`,
430 unless this is set to -1, in which case :mdp:`rlist` will be
435 Generate a pair list for groups of atoms, corresponding
436 to the charge groups in the topology. This option is no longer
445 Frequency to update the neighbor list. When dynamics and
446 :mdp:`verlet-buffer-tolerance` set, :mdp:`nstlist` is actually
447 a minimum value and :ref:`gmx mdrun` might increase it, unless
448 it is set to 1. With parallel simulations and/or non-bonded
449 force calculation on the GPU, a value of 20 or 40 often gives
450 the best performance.
454 The neighbor list is only constructed once and never
455 updated. This is mainly useful for vacuum simulations in which
456 all particles see each other. But vacuum simulations are
457 (temporarily) not supported.
467 Make a grid in the box and only check atoms in neighboring grid
468 cells when constructing a new neighbor list every
469 :mdp:`nstlist` steps. In large systems grid search is much
470 faster than simple search.
472 .. mdp-value:: simple
474 Check every atom in the box when constructing a new neighbor
475 list every :mdp:`nstlist` steps (only with :mdp-value:`cutoff-scheme=group`
482 Use periodic boundary conditions in all directions.
486 Use no periodic boundary conditions, ignore the box. To simulate
487 without cut-offs, set all cut-offs and :mdp:`nstlist` to 0. For
488 best performance without cut-offs on a single MPI rank, set
489 :mdp:`nstlist` to zero and :mdp-value:`ns-type=simple`.
493 Use periodic boundary conditions in x and y directions
494 only. This works only with :mdp-value:`ns-type=grid` and can be used
495 in combination with walls_. Without walls or with only one wall
496 the system size is infinite in the z direction. Therefore
497 pressure coupling or Ewald summation methods can not be
498 used. These disadvantages do not apply when two walls are used.
500 .. mdp:: periodic-molecules
504 molecules are finite, fast molecular PBC can be used
508 for systems with molecules that couple to themselves through the
509 periodic boundary conditions, this requires a slower PBC
510 algorithm and molecules are not made whole in the output
512 .. mdp:: verlet-buffer-tolerance
514 (0.005) [kJ mol\ :sup:`-1` ps\ :sup:`-1`]
516 Used when performing a simulation with dynamics. This sets
517 the maximum allowed error for pair interactions per particle caused
518 by the Verlet buffer, which indirectly sets :mdp:`rlist`. As both
519 :mdp:`nstlist` and the Verlet buffer size are fixed (for
520 performance reasons), particle pairs not in the pair list can
521 occasionally get within the cut-off distance during
522 :mdp:`nstlist` -1 steps. This causes very small jumps in the
523 energy. In a constant-temperature ensemble, these very small energy
524 jumps can be estimated for a given cut-off and :mdp:`rlist`. The
525 estimate assumes a homogeneous particle distribution, hence the
526 errors might be slightly underestimated for multi-phase
527 systems. (See the `reference manual`_ for details). For longer
528 pair-list life-time (:mdp:`nstlist` -1) * :mdp:`dt` the buffer is
529 overestimated, because the interactions between particles are
530 ignored. Combined with cancellation of errors, the actual drift of
531 the total energy is usually one to two orders of magnitude
532 smaller. Note that the generated buffer size takes into account
533 that the |Gromacs| pair-list setup leads to a reduction in the
534 drift by a factor 10, compared to a simple particle-pair based
535 list. Without dynamics (energy minimization etc.), the buffer is 5%
536 of the cut-off. For NVE simulations the initial temperature is
537 used, unless this is zero, in which case a buffer of 10% is
538 used. For NVE simulations the tolerance usually needs to be lowered
539 to achieve proper energy conservation on the nanosecond time
540 scale. To override the automated buffer setting, use
541 :mdp:`verlet-buffer-tolerance` =-1 and set :mdp:`rlist` manually.
546 Cut-off distance for the short-range neighbor list. With dynamics,
547 this is by default set by the :mdp:`verlet-buffer-tolerance` option
548 and the value of :mdp:`rlist` is ignored. Without dynamics, this
549 is by default set to the maximum cut-off plus 5% buffer, except
550 for test particle insertion, where the buffer is managed exactly
551 and automatically. For NVE simulations, where the automated
552 setting is not possible, the advised procedure is to run :ref:`gmx grompp`
553 with an NVT setup with the expected temperature and copy the resulting
554 value of :mdp:`rlist` to the NVE setup.
562 .. mdp-value:: Cut-off
564 Plain cut-off with pair list radius :mdp:`rlist` and
565 Coulomb cut-off :mdp:`rcoulomb`, where :mdp:`rlist` >=
570 Classical Ewald sum electrostatics. The real-space cut-off
571 :mdp:`rcoulomb` should be equal to :mdp:`rlist`. Use *e.g.*
572 :mdp:`rlist` =0.9, :mdp:`rcoulomb` =0.9. The highest magnitude
573 of wave vectors used in reciprocal space is controlled by
574 :mdp:`fourierspacing`. The relative accuracy of
575 direct/reciprocal space is controlled by :mdp:`ewald-rtol`.
577 NOTE: Ewald scales as O(N\ :sup:`3/2`) and is thus extremely slow for
578 large systems. It is included mainly for reference - in most
579 cases PME will perform much better.
583 Fast smooth Particle-Mesh Ewald (SPME) electrostatics. Direct
584 space is similar to the Ewald sum, while the reciprocal part is
585 performed with FFTs. Grid dimensions are controlled with
586 :mdp:`fourierspacing` and the interpolation order with
587 :mdp:`pme-order`. With a grid spacing of 0.1 nm and cubic
588 interpolation the electrostatic forces have an accuracy of
589 2-3*10\ :sup:`-4`. Since the error from the vdw-cutoff is larger than
590 this you might try 0.15 nm. When running in parallel the
591 interpolation parallelizes better than the FFT, so try
592 decreasing grid dimensions while increasing interpolation.
594 .. mdp-value:: P3M-AD
596 Particle-Particle Particle-Mesh algorithm with analytical
597 derivative for for long range electrostatic interactions. The
598 method and code is identical to SPME, except that the influence
599 function is optimized for the grid. This gives a slight increase
602 .. mdp-value:: Reaction-Field
604 Reaction field electrostatics with Coulomb cut-off
605 :mdp:`rcoulomb`, where :mdp:`rlist` >= :mdp:`rvdw`. The
606 dielectric constant beyond the cut-off is
607 :mdp:`epsilon-rf`. The dielectric constant can be set to
608 infinity by setting :mdp:`epsilon-rf` =0.
612 Currently unsupported.
613 :ref:`gmx mdrun` will now expect to find a file ``table.xvg``
614 with user-defined potential functions for repulsion, dispersion
615 and Coulomb. When pair interactions are present, :ref:`gmx
616 mdrun` also expects to find a file ``tablep.xvg`` for the pair
617 interactions. When the same interactions should be used for
618 non-bonded and pair interactions the user can specify the same
619 file name for both table files. These files should contain 7
620 columns: the ``x`` value, ``f(x)``, ``-f'(x)``, ``g(x)``,
621 ``-g'(x)``, ``h(x)``, ``-h'(x)``, where ``f(x)`` is the Coulomb
622 function, ``g(x)`` the dispersion function and ``h(x)`` the
623 repulsion function. When :mdp:`vdwtype` is not set to User the
624 values for ``g``, ``-g'``, ``h`` and ``-h'`` are ignored. For
625 the non-bonded interactions ``x`` values should run from 0 to
626 the largest cut-off distance + :mdp:`table-extension` and
627 should be uniformly spaced. For the pair interactions the table
628 length in the file will be used. The optimal spacing, which is
629 used for non-user tables, is ``0.002 nm`` when you run in mixed
630 precision or ``0.0005 nm`` when you run in double precision. The
631 function value at ``x=0`` is not important. More information is
632 in the printed manual.
634 .. mdp-value:: PME-Switch
636 Currently unsupported.
637 A combination of PME and a switch function for the direct-space
638 part (see above). :mdp:`rcoulomb` is allowed to be smaller than
641 .. mdp-value:: PME-User
643 Currently unsupported.
644 A combination of PME and user tables (see
645 above). :mdp:`rcoulomb` is allowed to be smaller than
646 :mdp:`rlist`. The PME mesh contribution is subtracted from the
647 user table by :ref:`gmx mdrun`. Because of this subtraction the
648 user tables should contain about 10 decimal places.
650 .. mdp-value:: PME-User-Switch
652 Currently unsupported.
653 A combination of PME-User and a switching function (see
654 above). The switching function is applied to final
655 particle-particle interaction, *i.e.* both to the user supplied
656 function and the PME Mesh correction part.
658 .. mdp:: coulomb-modifier
660 .. mdp-value:: Potential-shift
662 Shift the Coulomb potential by a constant such that it is zero
663 at the cut-off. This makes the potential the integral of the
664 force. Note that this does not affect the forces or the
669 Use an unmodified Coulomb potential. This can be useful
670 when comparing energies with those computed with other software.
672 .. mdp:: rcoulomb-switch
675 where to start switching the Coulomb potential, only relevant
676 when force or potential switching is used
681 The distance for the Coulomb cut-off. Note that with PME this value
682 can be increased by the PME tuning in :ref:`gmx mdrun` along with
683 the PME grid spacing.
688 The relative dielectric constant. A value of 0 means infinity.
693 The relative dielectric constant of the reaction field. This
694 is only used with reaction-field electrostatics. A value of 0
703 .. mdp-value:: Cut-off
705 Plain cut-off with pair list radius :mdp:`rlist` and VdW
706 cut-off :mdp:`rvdw`, where :mdp:`rlist` >= :mdp:`rvdw`.
710 Fast smooth Particle-mesh Ewald (SPME) for VdW interactions. The
711 grid dimensions are controlled with :mdp:`fourierspacing` in
712 the same way as for electrostatics, and the interpolation order
713 is controlled with :mdp:`pme-order`. The relative accuracy of
714 direct/reciprocal space is controlled by :mdp:`ewald-rtol-lj`,
715 and the specific combination rules that are to be used by the
716 reciprocal routine are set using :mdp:`lj-pme-comb-rule`.
720 This functionality is deprecated and replaced by using
721 :mdp-value:`vdwtype=Cut-off` with :mdp-value:`vdw-modifier=Force-switch`.
722 The LJ (not Buckingham) potential is decreased over the whole range and
723 the forces decay smoothly to zero between :mdp:`rvdw-switch` and
726 .. mdp-value:: Switch
728 This functionality is deprecated and replaced by using
729 :mdp-value:`vdwtype=Cut-off` with :mdp-value:`vdw-modifier=Potential-switch`.
730 The LJ (not Buckingham) potential is normal out to :mdp:`rvdw-switch`, after
731 which it is switched off to reach zero at :mdp:`rvdw`. Both the
732 potential and force functions are continuously smooth, but be
733 aware that all switch functions will give rise to a bulge
734 (increase) in the force (since we are switching the
739 Currently unsupported.
740 See user for :mdp:`coulombtype`. The function value at zero is
741 not important. When you want to use LJ correction, make sure
742 that :mdp:`rvdw` corresponds to the cut-off in the user-defined
743 function. When :mdp:`coulombtype` is not set to User the values
744 for the ``f`` and ``-f'`` columns are ignored.
746 .. mdp:: vdw-modifier
748 .. mdp-value:: Potential-shift
750 Shift the Van der Waals potential by a constant such that it is
751 zero at the cut-off. This makes the potential the integral of
752 the force. Note that this does not affect the forces or the
757 Use an unmodified Van der Waals potential. This can be useful
758 when comparing energies with those computed with other software.
760 .. mdp-value:: Force-switch
762 Smoothly switches the forces to zero between :mdp:`rvdw-switch`
763 and :mdp:`rvdw`. This shifts the potential shift over the whole
764 range and switches it to zero at the cut-off. Note that this is
765 more expensive to calculate than a plain cut-off and it is not
766 required for energy conservation, since Potential-shift
767 conserves energy just as well.
769 .. mdp-value:: Potential-switch
771 Smoothly switches the potential to zero between
772 :mdp:`rvdw-switch` and :mdp:`rvdw`. Note that this introduces
773 articifically large forces in the switching region and is much
774 more expensive to calculate. This option should only be used if
775 the force field you are using requires this.
780 where to start switching the LJ force and possibly the potential,
781 only relevant when force or potential switching is used
786 distance for the LJ or Buckingham cut-off
792 don't apply any correction
794 .. mdp-value:: EnerPres
796 apply long range dispersion corrections for Energy and Pressure
800 apply long range dispersion corrections for Energy only
806 .. mdp:: table-extension
809 Extension of the non-bonded potential lookup tables beyond the
810 largest cut-off distance. With actual non-bonded interactions
811 the tables are never accessed beyond the cut-off. But a longer
812 table length might be needed for the 1-4 interactions, which
813 are always tabulated irrespective of the use of tables for
814 the non-bonded interactions.
816 .. mdp:: energygrp-table
818 Currently unsupported.
819 When user tables are used for electrostatics and/or VdW, here one
820 can give pairs of energy groups for which seperate user tables
821 should be used. The two energy groups will be appended to the table
822 file name, in order of their definition in :mdp:`energygrps`,
823 seperated by underscores. For example, if ``energygrps = Na Cl
824 Sol`` and ``energygrp-table = Na Na Na Cl``, :ref:`gmx mdrun` will
825 read ``table_Na_Na.xvg`` and ``table_Na_Cl.xvg`` in addition to the
826 normal ``table.xvg`` which will be used for all other energy group
833 .. mdp:: fourierspacing
836 For ordinary Ewald, the ratio of the box dimensions and the spacing
837 determines a lower bound for the number of wave vectors to use in
838 each (signed) direction. For PME and P3M, that ratio determines a
839 lower bound for the number of Fourier-space grid points that will
840 be used along that axis. In all cases, the number for each
841 direction can be overridden by entering a non-zero value for that
842 :mdp:`fourier-nx` direction. For optimizing the relative load of
843 the particle-particle interactions and the mesh part of PME, it is
844 useful to know that the accuracy of the electrostatics remains
845 nearly constant when the Coulomb cut-off and the PME grid spacing
846 are scaled by the same factor. Note that this spacing can be scaled
847 up along with :mdp:`rcoulomb` by the PME tuning in :ref:`gmx mdrun`.
854 Highest magnitude of wave vectors in reciprocal space when using Ewald.
855 Grid size when using PME or P3M. These values override
856 :mdp:`fourierspacing` per direction. The best choice is powers of
857 2, 3, 5 and 7. Avoid large primes. Note that these grid sizes can
858 be reduced along with scaling up :mdp:`rcoulomb` by the PME tuning
864 Interpolation order for PME. 4 equals cubic interpolation. You
865 might try 6/8/10 when running in parallel and simultaneously
866 decrease grid dimension.
871 The relative strength of the Ewald-shifted direct potential at
872 :mdp:`rcoulomb` is given by :mdp:`ewald-rtol`. Decreasing this
873 will give a more accurate direct sum, but then you need more wave
874 vectors for the reciprocal sum.
876 .. mdp:: ewald-rtol-lj
879 When doing PME for VdW-interactions, :mdp:`ewald-rtol-lj` is used
880 to control the relative strength of the dispersion potential at
881 :mdp:`rvdw` in the same way as :mdp:`ewald-rtol` controls the
882 electrostatic potential.
884 .. mdp:: lj-pme-comb-rule
887 The combination rules used to combine VdW-parameters in the
888 reciprocal part of LJ-PME. Geometric rules are much faster than
889 Lorentz-Berthelot and usually the recommended choice, even when the
890 rest of the force field uses the Lorentz-Berthelot rules.
892 .. mdp-value:: Geometric
894 Apply geometric combination rules
896 .. mdp-value:: Lorentz-Berthelot
898 Apply Lorentz-Berthelot combination rules
900 .. mdp:: ewald-geometry
904 The Ewald sum is performed in all three dimensions.
908 The reciprocal sum is still performed in 3D, but a force and
909 potential correction applied in the `z` dimension to produce a
910 pseudo-2D summation. If your system has a slab geometry in the
911 `x-y` plane you can try to increase the `z`-dimension of the box
912 (a box height of 3 times the slab height is usually ok) and use
915 .. mdp:: epsilon-surface
918 This controls the dipole correction to the Ewald summation in
919 3D. The default value of zero means it is turned off. Turn it on by
920 setting it to the value of the relative permittivity of the
921 imaginary surface around your infinite system. Be careful - you
922 shouldn't use this if you have free mobile charges in your
923 system. This value does not affect the slab 3DC variant of the long
934 No temperature coupling.
936 .. mdp-value:: berendsen
938 Temperature coupling with a Berendsen thermostat to a bath with
939 temperature :mdp:`ref-t`, with time constant
940 :mdp:`tau-t`. Several groups can be coupled separately, these
941 are specified in the :mdp:`tc-grps` field separated by spaces.
943 .. mdp-value:: nose-hoover
945 Temperature coupling using a Nose-Hoover extended ensemble. The
946 reference temperature and coupling groups are selected as above,
947 but in this case :mdp:`tau-t` controls the period of the
948 temperature fluctuations at equilibrium, which is slightly
949 different from a relaxation time. For NVT simulations the
950 conserved energy quantity is written to the energy and log files.
952 .. mdp-value:: andersen
954 Temperature coupling by randomizing a fraction of the particle velocities
955 at each timestep. Reference temperature and coupling groups are
956 selected as above. :mdp:`tau-t` is the average time between
957 randomization of each molecule. Inhibits particle dynamics
958 somewhat, but little or no ergodicity issues. Currently only
959 implemented with velocity Verlet, and not implemented with
962 .. mdp-value:: andersen-massive
964 Temperature coupling by randomizing velocities of all particles at
965 infrequent timesteps. Reference temperature and coupling groups are
966 selected as above. :mdp:`tau-t` is the time between
967 randomization of all molecules. Inhibits particle dynamics
968 somewhat, but little or no ergodicity issues. Currently only
969 implemented with velocity Verlet.
971 .. mdp-value:: v-rescale
973 Temperature coupling using velocity rescaling with a stochastic
974 term (JCP 126, 014101). This thermostat is similar to Berendsen
975 coupling, with the same scaling using :mdp:`tau-t`, but the
976 stochastic term ensures that a proper canonical ensemble is
977 generated. The random seed is set with :mdp:`ld-seed`. This
978 thermostat works correctly even for :mdp:`tau-t` =0. For NVT
979 simulations the conserved energy quantity is written to the
985 The frequency for coupling the temperature. The default value of -1
986 sets :mdp:`nsttcouple` equal to :mdp:`nstlist`, unless
987 :mdp:`nstlist` <=0, then a value of 10 is used. For velocity
988 Verlet integrators :mdp:`nsttcouple` is set to 1.
990 .. mdp:: nh-chain-length
993 The number of chained Nose-Hoover thermostats for velocity Verlet
994 integrators, the leap-frog :mdp-value:`integrator=md` integrator
995 only supports 1. Data for the NH chain variables is not printed
996 to the :ref:`edr` file by default, but can be turned on with the
997 :mdp:`print-nose-hoover-chain-variables` option.
999 .. mdp:: print-nose-hoover-chain-variables
1003 Do not store Nose-Hoover chain variables in the energy file.
1007 Store all positions and velocities of the Nose-Hoover chain
1012 groups to couple to separate temperature baths
1017 time constant for coupling (one for each group in
1018 :mdp:`tc-grps`), -1 means no temperature coupling
1023 reference temperature for coupling (one for each group in
1034 No pressure coupling. This means a fixed box size.
1036 .. mdp-value:: Berendsen
1038 Exponential relaxation pressure coupling with time constant
1039 :mdp:`tau-p`. The box is scaled every :mdp:`nstpcouple` steps. It has been
1040 argued that this does not yield a correct thermodynamic
1041 ensemble, but it is the most efficient way to scale a box at the
1044 .. mdp-value:: Parrinello-Rahman
1046 Extended-ensemble pressure coupling where the box vectors are
1047 subject to an equation of motion. The equation of motion for the
1048 atoms is coupled to this. No instantaneous scaling takes
1049 place. As for Nose-Hoover temperature coupling the time constant
1050 :mdp:`tau-p` is the period of pressure fluctuations at
1051 equilibrium. This is probably a better method when you want to
1052 apply pressure scaling during data collection, but beware that
1053 you can get very large oscillations if you are starting from a
1054 different pressure. For simulations where the exact fluctations
1055 of the NPT ensemble are important, or if the pressure coupling
1056 time is very short it may not be appropriate, as the previous
1057 time step pressure is used in some steps of the |Gromacs|
1058 implementation for the current time step pressure.
1062 Martyna-Tuckerman-Tobias-Klein implementation, only useable with
1063 :mdp-value:`integrator=md-vv` or :mdp-value:`integrator=md-vv-avek`, very similar to
1064 Parrinello-Rahman. As for Nose-Hoover temperature coupling the
1065 time constant :mdp:`tau-p` is the period of pressure
1066 fluctuations at equilibrium. This is probably a better method
1067 when you want to apply pressure scaling during data collection,
1068 but beware that you can get very large oscillations if you are
1069 starting from a different pressure. Currently (as of version
1070 5.1), it only supports isotropic scaling, and only works without
1075 Specifies the kind of isotropy of the pressure coupling used. Each
1076 kind takes one or more values for :mdp:`compressibility` and
1077 :mdp:`ref-p`. Only a single value is permitted for :mdp:`tau-p`.
1079 .. mdp-value:: isotropic
1081 Isotropic pressure coupling with time constant
1082 :mdp:`tau-p`. One value each for :mdp:`compressibility` and
1083 :mdp:`ref-p` is required.
1085 .. mdp-value:: semiisotropic
1087 Pressure coupling which is isotropic in the ``x`` and ``y``
1088 direction, but different in the ``z`` direction. This can be
1089 useful for membrane simulations. Two values each for
1090 :mdp:`compressibility` and :mdp:`ref-p` are required, for
1091 ``x/y`` and ``z`` directions respectively.
1093 .. mdp-value:: anisotropic
1095 Same as before, but 6 values are needed for ``xx``, ``yy``, ``zz``,
1096 ``xy/yx``, ``xz/zx`` and ``yz/zy`` components,
1097 respectively. When the off-diagonal compressibilities are set to
1098 zero, a rectangular box will stay rectangular. Beware that
1099 anisotropic scaling can lead to extreme deformation of the
1102 .. mdp-value:: surface-tension
1104 Surface tension coupling for surfaces parallel to the
1105 xy-plane. Uses normal pressure coupling for the `z`-direction,
1106 while the surface tension is coupled to the `x/y` dimensions of
1107 the box. The first :mdp:`ref-p` value is the reference surface
1108 tension times the number of surfaces ``bar nm``, the second
1109 value is the reference `z`-pressure ``bar``. The two
1110 :mdp:`compressibility` values are the compressibility in the
1111 `x/y` and `z` direction respectively. The value for the
1112 `z`-compressibility should be reasonably accurate since it
1113 influences the convergence of the surface-tension, it can also
1114 be set to zero to have a box with constant height.
1119 The frequency for coupling the pressure. The default value of -1
1120 sets :mdp:`nstpcouple` equal to :mdp:`nstlist`, unless
1121 :mdp:`nstlist` <=0, then a value of 10 is used. For velocity
1122 Verlet integrators :mdp:`nstpcouple` is set to 1.
1127 The time constant for pressure coupling (one value for all
1130 .. mdp:: compressibility
1133 The compressibility (NOTE: this is now really in bar\ :sup:`-1`) For water at 1
1134 atm and 300 K the compressibility is 4.5e-5 bar\ :sup:`-1`. The number of
1135 required values is implied by :mdp:`pcoupltype`.
1140 The reference pressure for coupling. The number of required values
1141 is implied by :mdp:`pcoupltype`.
1143 .. mdp:: refcoord-scaling
1147 The reference coordinates for position restraints are not
1148 modified. Note that with this option the virial and pressure
1149 might be ill defined, see :ref:`here <reference-manual-position-restraints>`
1154 The reference coordinates are scaled with the scaling matrix of
1155 the pressure coupling.
1159 Scale the center of mass of the reference coordinates with the
1160 scaling matrix of the pressure coupling. The vectors of each
1161 reference coordinate to the center of mass are not scaled. Only
1162 one COM is used, even when there are multiple molecules with
1163 position restraints. For calculating the COM of the reference
1164 coordinates in the starting configuration, periodic boundary
1165 conditions are not taken into account. Note that with this option
1166 the virial and pressure might be ill defined, see
1167 :ref:`here <reference-manual-position-restraints>` for more details.
1173 Simulated annealing is controlled separately for each temperature
1174 group in |Gromacs|. The reference temperature is a piecewise linear
1175 function, but you can use an arbitrary number of points for each
1176 group, and choose either a single sequence or a periodic behaviour for
1177 each group. The actual annealing is performed by dynamically changing
1178 the reference temperature used in the thermostat algorithm selected,
1179 so remember that the system will usually not instantaneously reach the
1180 reference temperature!
1184 Type of annealing for each temperature group
1188 No simulated annealing - just couple to reference temperature value.
1190 .. mdp-value:: single
1192 A single sequence of annealing points. If your simulation is
1193 longer than the time of the last point, the temperature will be
1194 coupled to this constant value after the annealing sequence has
1195 reached the last time point.
1197 .. mdp-value:: periodic
1199 The annealing will start over at the first reference point once
1200 the last reference time is reached. This is repeated until the
1203 .. mdp:: annealing-npoints
1205 A list with the number of annealing reference/control points used
1206 for each temperature group. Use 0 for groups that are not
1207 annealed. The number of entries should equal the number of
1210 .. mdp:: annealing-time
1212 List of times at the annealing reference/control points for each
1213 group. If you are using periodic annealing, the times will be used
1214 modulo the last value, *i.e.* if the values are 0, 5, 10, and 15,
1215 the coupling will restart at the 0ps value after 15ps, 30ps, 45ps,
1216 etc. The number of entries should equal the sum of the numbers
1217 given in :mdp:`annealing-npoints`.
1219 .. mdp:: annealing-temp
1221 List of temperatures at the annealing reference/control points for
1222 each group. The number of entries should equal the sum of the
1223 numbers given in :mdp:`annealing-npoints`.
1225 Confused? OK, let's use an example. Assume you have two temperature
1226 groups, set the group selections to ``annealing = single periodic``,
1227 the number of points of each group to ``annealing-npoints = 3 4``, the
1228 times to ``annealing-time = 0 3 6 0 2 4 6`` and finally temperatures
1229 to ``annealing-temp = 298 280 270 298 320 320 298``. The first group
1230 will be coupled to 298K at 0ps, but the reference temperature will
1231 drop linearly to reach 280K at 3ps, and then linearly between 280K and
1232 270K from 3ps to 6ps. After this is stays constant, at 270K. The
1233 second group is coupled to 298K at 0ps, it increases linearly to 320K
1234 at 2ps, where it stays constant until 4ps. Between 4ps and 6ps it
1235 decreases to 298K, and then it starts over with the same pattern
1236 again, *i.e.* rising linearly from 298K to 320K between 6ps and
1237 8ps. Check the summary printed by :ref:`gmx grompp` if you are unsure!
1247 Do not generate velocities. The velocities are set to zero
1248 when there are no velocities in the input structure file.
1252 Generate velocities in :ref:`gmx grompp` according to a
1253 Maxwell distribution at temperature :mdp:`gen-temp`, with
1254 random seed :mdp:`gen-seed`. This is only meaningful with
1255 :mdp-value:`integrator=md`.
1260 temperature for Maxwell distribution
1265 used to initialize random generator for random velocities,
1266 when :mdp:`gen-seed` is set to -1, a pseudo random seed is
1273 .. mdp:: constraints
1275 Controls which bonds in the topology will be converted to rigid
1276 holonomic constraints. Note that typical rigid water models do not
1277 have bonds, but rather a specialized ``[settles]`` directive, so
1278 are not affected by this keyword.
1282 No bonds converted to constraints.
1284 .. mdp-value:: h-bonds
1286 Convert the bonds with H-atoms to constraints.
1288 .. mdp-value:: all-bonds
1290 Convert all bonds to constraints.
1292 .. mdp-value:: h-angles
1294 Convert all bonds to constraints and convert the angles that
1295 involve H-atoms to bond-constraints.
1297 .. mdp-value:: all-angles
1299 Convert all bonds to constraints and all angles to bond-constraints.
1301 .. mdp:: constraint-algorithm
1303 Chooses which solver satisfies any non-SETTLE holonomic
1306 .. mdp-value:: LINCS
1308 LINear Constraint Solver. With domain decomposition the parallel
1309 version P-LINCS is used. The accuracy in set with
1310 :mdp:`lincs-order`, which sets the number of matrices in the
1311 expansion for the matrix inversion. After the matrix inversion
1312 correction the algorithm does an iterative correction to
1313 compensate for lengthening due to rotation. The number of such
1314 iterations can be controlled with :mdp:`lincs-iter`. The root
1315 mean square relative constraint deviation is printed to the log
1316 file every :mdp:`nstlog` steps. If a bond rotates more than
1317 :mdp:`lincs-warnangle` in one step, a warning will be printed
1318 both to the log file and to ``stderr``. LINCS should not be used
1319 with coupled angle constraints.
1321 .. mdp-value:: SHAKE
1323 SHAKE is slightly slower and less stable than LINCS, but does
1324 work with angle constraints. The relative tolerance is set with
1325 :mdp:`shake-tol`, 0.0001 is a good value for "normal" MD. SHAKE
1326 does not support constraints between atoms on different
1327 decomposition domains, so it can only be used with domain
1328 decomposition when so-called update-groups are used, which is
1329 usally the case when only bonds involving hydrogens are
1330 constrained. SHAKE can not be used with energy minimization.
1332 .. mdp:: continuation
1334 This option was formerly known as ``unconstrained-start``.
1338 apply constraints to the start configuration and reset shells
1342 do not apply constraints to the start configuration and do not
1343 reset shells, useful for exact coninuation and reruns
1348 relative tolerance for SHAKE
1350 .. mdp:: lincs-order
1353 Highest order in the expansion of the constraint coupling
1354 matrix. When constraints form triangles, an additional expansion of
1355 the same order is applied on top of the normal expansion only for
1356 the couplings within such triangles. For "normal" MD simulations an
1357 order of 4 usually suffices, 6 is needed for large time-steps with
1358 virtual sites or BD. For accurate energy minimization an order of 8
1359 or more might be required. With domain decomposition, the cell size
1360 is limited by the distance spanned by :mdp:`lincs-order` +1
1361 constraints. When one wants to scale further than this limit, one
1362 can decrease :mdp:`lincs-order` and increase :mdp:`lincs-iter`,
1363 since the accuracy does not deteriorate when (1+ :mdp:`lincs-iter`
1364 )* :mdp:`lincs-order` remains constant.
1369 Number of iterations to correct for rotational lengthening in
1370 LINCS. For normal runs a single step is sufficient, but for NVE
1371 runs where you want to conserve energy accurately or for accurate
1372 energy minimization you might want to increase it to 2.
1374 .. mdp:: lincs-warnangle
1377 maximum angle that a bond can rotate before LINCS will complain
1383 bonds are represented by a harmonic potential
1387 bonds are represented by a Morse potential
1390 Energy group exclusions
1391 ^^^^^^^^^^^^^^^^^^^^^^^
1393 .. mdp:: energygrp-excl
1395 Pairs of energy groups for which all non-bonded interactions are
1396 excluded. An example: if you have two energy groups ``Protein`` and
1397 ``SOL``, specifying ``energygrp-excl = Protein Protein SOL SOL``
1398 would give only the non-bonded interactions between the protein and
1399 the solvent. This is especially useful for speeding up energy
1400 calculations with ``mdrun -rerun`` and for excluding interactions
1401 within frozen groups.
1410 When set to 1 there is a wall at ``z=0``, when set to 2 there is
1411 also a wall at ``z=z-box``. Walls can only be used with :mdp:`pbc`
1412 ``=xy``. When set to 2, pressure coupling and Ewald summation can be
1413 used (it is usually best to use semiisotropic pressure coupling
1414 with the ``x/y`` compressibility set to 0, as otherwise the surface
1415 area will change). Walls interact wit the rest of the system
1416 through an optional :mdp:`wall-atomtype`. Energy groups ``wall0``
1417 and ``wall1`` (for :mdp:`nwall` =2) are added automatically to
1418 monitor the interaction of energy groups with each wall. The center
1419 of mass motion removal will be turned off in the ``z``-direction.
1421 .. mdp:: wall-atomtype
1423 the atom type name in the force field for each wall. By (for
1424 example) defining a special wall atom type in the topology with its
1425 own combination rules, this allows for independent tuning of the
1426 interaction of each atomtype with the walls.
1432 LJ integrated over the volume behind the wall: 9-3 potential
1436 LJ integrated over the wall surface: 10-4 potential
1440 direct LJ potential with the ``z`` distance from the wall
1444 user defined potentials indexed with the ``z`` distance from the
1445 wall, the tables are read analogously to the
1446 :mdp:`energygrp-table` option, where the first name is for a
1447 "normal" energy group and the second name is ``wall0`` or
1448 ``wall1``, only the dispersion and repulsion columns are used
1450 .. mdp:: wall-r-linpot
1453 Below this distance from the wall the potential is continued
1454 linearly and thus the force is constant. Setting this option to a
1455 postive value is especially useful for equilibration when some
1456 atoms are beyond a wall. When the value is <=0 (<0 for
1457 :mdp:`wall-type` =table), a fatal error is generated when atoms
1460 .. mdp:: wall-density
1462 [nm\ :sup:`-3`] / [nm\ :sup:`-2`]
1463 the number density of the atoms for each wall for wall types 9-3
1466 .. mdp:: wall-ewald-zfac
1469 The scaling factor for the third box vector for Ewald summation
1470 only, the minimum is 2. Ewald summation can only be used with
1471 :mdp:`nwall` =2, where one should use :mdp:`ewald-geometry`
1472 ``=3dc``. The empty layer in the box serves to decrease the
1473 unphysical Coulomb interaction between periodic images.
1479 Note that where pulling coordinates are applicable, there can be more
1480 than one (set with :mdp:`pull-ncoords`) and multiple related :ref:`mdp`
1481 variables will exist accordingly. Documentation references to things
1482 like :mdp:`pull-coord1-vec` should be understood to apply to to the
1483 applicable pulling coordinate, eg. the second pull coordinate is described by
1484 pull-coord2-vec, pull-coord2-k, and so on.
1490 No center of mass pulling. All the following pull options will
1491 be ignored (and if present in the :ref:`mdp` file, they unfortunately
1496 Center of mass pulling will be applied on 1 or more groups using
1497 1 or more pull coordinates.
1499 .. mdp:: pull-cylinder-r
1502 the radius of the cylinder for :mdp-value:`pull-coord1-geometry=cylinder`
1504 .. mdp:: pull-constr-tol
1507 the relative constraint tolerance for constraint pulling
1509 .. mdp:: pull-print-com
1513 do not print the COM for any group
1517 print the COM of all groups for all pull coordinates
1519 .. mdp:: pull-print-ref-value
1523 do not print the reference value for each pull coordinate
1527 print the reference value for each pull coordinate
1529 .. mdp:: pull-print-components
1533 only print the distance for each pull coordinate
1537 print the distance and Cartesian components selected in
1538 :mdp:`pull-coord1-dim`
1540 .. mdp:: pull-nstxout
1543 frequency for writing out the COMs of all the pull group (0 is
1546 .. mdp:: pull-nstfout
1549 frequency for writing out the force of all the pulled group
1552 .. mdp:: pull-pbc-ref-prev-step-com
1556 Use the reference atom (:mdp:`pull-group1-pbcatom`) for the
1557 treatment of periodic boundary conditions.
1561 Use the COM of the previous step as reference for the treatment
1562 of periodic boundary conditions. The reference is initialized
1563 using the reference atom (:mdp:`pull-group1-pbcatom`), which should
1564 be located centrally in the group. Using the COM from the
1565 previous step can be useful if one or more pull groups are large.
1567 .. mdp:: pull-xout-average
1571 Write the instantaneous coordinates for all the pulled groups.
1575 Write the average coordinates (since last output) for all the
1576 pulled groups. N.b., some analysis tools might expect instantaneous
1579 .. mdp:: pull-fout-average
1583 Write the instantaneous force for all the pulled groups.
1587 Write the average force (since last output) for all the
1588 pulled groups. N.b., some analysis tools might expect instantaneous
1591 .. mdp:: pull-ngroups
1594 The number of pull groups, not including the absolute reference
1595 group, when used. Pull groups can be reused in multiple pull
1596 coordinates. Below only the pull options for group 1 are given,
1597 further groups simply increase the group index number.
1599 .. mdp:: pull-ncoords
1602 The number of pull coordinates. Below only the pull options for
1603 coordinate 1 are given, further coordinates simply increase the
1604 coordinate index number.
1606 .. mdp:: pull-group1-name
1608 The name of the pull group, is looked up in the index file or in
1609 the default groups to obtain the atoms involved.
1611 .. mdp:: pull-group1-weights
1613 Optional relative weights which are multiplied with the masses of
1614 the atoms to give the total weight for the COM. The number should
1615 be 0, meaning all 1, or the number of atoms in the pull group.
1617 .. mdp:: pull-group1-pbcatom
1620 The reference atom for the treatment of periodic boundary
1621 conditions inside the group (this has no effect on the treatment of
1622 the pbc between groups). This option is only important when the
1623 diameter of the pull group is larger than half the shortest box
1624 vector. For determining the COM, all atoms in the group are put at
1625 their periodic image which is closest to
1626 :mdp:`pull-group1-pbcatom`. A value of 0 means that the middle
1627 atom (number wise) is used, which is only safe for small groups.
1628 :ref:`gmx grompp` checks that the maximum distance from the reference
1629 atom (specifically chosen, or not) to the other atoms in the group
1630 is not too large. This parameter is not used with
1631 :mdp:`pull-coord1-geometry` cylinder. A value of -1 turns on cosine
1632 weighting, which is useful for a group of molecules in a periodic
1633 system, *e.g.* a water slab (see Engin et al. J. Chem. Phys. B
1636 .. mdp:: pull-coord1-type
1638 .. mdp-value:: umbrella
1640 Center of mass pulling using an umbrella potential between the
1641 reference group and one or more groups.
1643 .. mdp-value:: constraint
1645 Center of mass pulling using a constraint between the reference
1646 group and one or more groups. The setup is identical to the
1647 option umbrella, except for the fact that a rigid constraint is
1648 applied instead of a harmonic potential.
1650 .. mdp-value:: constant-force
1652 Center of mass pulling using a linear potential and therefore a
1653 constant force. For this option there is no reference position
1654 and therefore the parameters :mdp:`pull-coord1-init` and
1655 :mdp:`pull-coord1-rate` are not used.
1657 .. mdp-value:: flat-bottom
1659 At distances above :mdp:`pull-coord1-init` a harmonic potential
1660 is applied, otherwise no potential is applied.
1662 .. mdp-value:: flat-bottom-high
1664 At distances below :mdp:`pull-coord1-init` a harmonic potential
1665 is applied, otherwise no potential is applied.
1667 .. mdp-value:: external-potential
1669 An external potential that needs to be provided by another
1672 .. mdp:: pull-coord1-potential-provider
1674 The name of the external module that provides the potential for
1675 the case where :mdp:`pull-coord1-type` is external-potential.
1677 .. mdp:: pull-coord1-geometry
1679 .. mdp-value:: distance
1681 Pull along the vector connecting the two groups. Components can
1682 be selected with :mdp:`pull-coord1-dim`.
1684 .. mdp-value:: direction
1686 Pull in the direction of :mdp:`pull-coord1-vec`.
1688 .. mdp-value:: direction-periodic
1690 As :mdp-value:`pull-coord1-geometry=direction`, but allows the distance to be larger
1691 than half the box size. With this geometry the box should not be
1692 dynamic (*e.g.* no pressure scaling) in the pull dimensions and
1693 the pull force is not added to virial.
1695 .. mdp-value:: direction-relative
1697 As :mdp-value:`pull-coord1-geometry=direction`, but the pull vector is the vector
1698 that points from the COM of a third to the COM of a fourth pull
1699 group. This means that 4 groups need to be supplied in
1700 :mdp:`pull-coord1-groups`. Note that the pull force will give
1701 rise to a torque on the pull vector, which is turn leads to
1702 forces perpendicular to the pull vector on the two groups
1703 defining the vector. If you want a pull group to move between
1704 the two groups defining the vector, simply use the union of
1705 these two groups as the reference group.
1707 .. mdp-value:: cylinder
1709 Designed for pulling with respect to a layer where the reference
1710 COM is given by a local cylindrical part of the reference group.
1711 The pulling is in the direction of :mdp:`pull-coord1-vec`. From
1712 the first of the two groups in :mdp:`pull-coord1-groups` a
1713 cylinder is selected around the axis going through the COM of
1714 the second group with direction :mdp:`pull-coord1-vec` with
1715 radius :mdp:`pull-cylinder-r`. Weights of the atoms decrease
1716 continously to zero as the radial distance goes from 0 to
1717 :mdp:`pull-cylinder-r` (mass weighting is also used). The radial
1718 dependence gives rise to radial forces on both pull groups.
1719 Note that the radius should be smaller than half the box size.
1720 For tilted cylinders they should be even smaller than half the
1721 box size since the distance of an atom in the reference group
1722 from the COM of the pull group has both a radial and an axial
1723 component. This geometry is not supported with constraint
1726 .. mdp-value:: angle
1728 Pull along an angle defined by four groups. The angle is
1729 defined as the angle between two vectors: the vector connecting
1730 the COM of the first group to the COM of the second group and
1731 the vector connecting the COM of the third group to the COM of
1734 .. mdp-value:: angle-axis
1736 As :mdp-value:`pull-coord1-geometry=angle` but the second vector is given by :mdp:`pull-coord1-vec`.
1737 Thus, only the two groups that define the first vector need to be given.
1739 .. mdp-value:: dihedral
1741 Pull along a dihedral angle defined by six groups. These pairwise
1742 define three vectors: the vector connecting the COM of group 1
1743 to the COM of group 2, the COM of group 3 to the COM of group 4,
1744 and the COM of group 5 to the COM group 6. The dihedral angle is
1745 then defined as the angle between two planes: the plane spanned by the
1746 the two first vectors and the plane spanned the two last vectors.
1749 .. mdp:: pull-coord1-groups
1751 The group indices on which this pull coordinate will operate.
1752 The number of group indices required is geometry dependent.
1753 The first index can be 0, in which case an
1754 absolute reference of :mdp:`pull-coord1-origin` is used. With an
1755 absolute reference the system is no longer translation invariant
1756 and one should think about what to do with the center of mass
1759 .. mdp:: pull-coord1-dim
1762 Selects the dimensions that this pull coordinate acts on and that
1763 are printed to the output files when
1764 :mdp:`pull-print-components` = :mdp-value:`pull-coord1-start=yes`. With
1765 :mdp:`pull-coord1-geometry` = :mdp-value:`pull-coord1-geometry=distance`, only Cartesian
1766 components set to Y contribute to the distance. Thus setting this
1767 to Y Y N results in a distance in the x/y plane. With other
1768 geometries all dimensions with non-zero entries in
1769 :mdp:`pull-coord1-vec` should be set to Y, the values for other
1770 dimensions only affect the output.
1772 .. mdp:: pull-coord1-origin
1775 The pull reference position for use with an absolute reference.
1777 .. mdp:: pull-coord1-vec
1780 The pull direction. :ref:`gmx grompp` normalizes the vector.
1782 .. mdp:: pull-coord1-start
1786 do not modify :mdp:`pull-coord1-init`
1790 add the COM distance of the starting conformation to
1791 :mdp:`pull-coord1-init`
1793 .. mdp:: pull-coord1-init
1796 The reference distance or reference angle at t=0.
1798 .. mdp:: pull-coord1-rate
1800 (0) [nm/ps] or [deg/ps]
1801 The rate of change of the reference position or reference angle.
1803 .. mdp:: pull-coord1-k
1805 (0) [kJ mol\ :sup:`-1` nm\ :sup:`-2`] or [kJ mol\ :sup:`-1` nm\ :sup:`-1`] or
1806 [kJ mol\ :sup:`-1` rad\ :sup:`-2`] or [kJ mol\ :sup:`-1` rad\ :sup:`-1`]
1807 The force constant. For umbrella pulling this is the harmonic force
1808 constant in kJ mol\ :sup:`-1` nm\ :sup:`-2` (or kJ mol\ :sup:`-1` rad\ :sup:`-2`
1809 for angles). For constant force pulling this is the
1810 force constant of the linear potential, and thus the negative (!)
1811 of the constant force in kJ mol\ :sup:`-1` nm\ :sup:`-1`
1812 (or kJ mol\ :sup:`-1` rad\ :sup:`-1` for angles).
1813 Note that for angles the force constant is expressed in terms of radians
1814 (while :mdp:`pull-coord1-init` and :mdp:`pull-coord1-rate` are expressed in degrees).
1816 .. mdp:: pull-coord1-kB
1818 (pull-k1) [kJ mol\ :sup:`-1` nm\ :sup:`-2`] or [kJ mol\ :sup:`-1` nm\ :sup:`-1`]
1819 or [kJ mol\ :sup:`-1` rad\ :sup:`-2`] or [kJ mol\ :sup:`-1` rad\ :sup:`-1`]
1820 As :mdp:`pull-coord1-k`, but for state B. This is only used when
1821 :mdp:`free-energy` is turned on. The force constant is then (1 -
1822 lambda) * :mdp:`pull-coord1-k` + lambda * :mdp:`pull-coord1-kB`.
1824 AWH adaptive biasing
1825 ^^^^^^^^^^^^^^^^^^^^
1835 Adaptively bias a reaction coordinate using the AWH method and estimate
1836 the corresponding PMF. The PMF and other AWH data are written to energy
1837 file at an interval set by :mdp:`awh-nstout` and can be extracted with
1838 the ``gmx awh`` tool. The AWH coordinate can be
1839 multidimensional and is defined by mapping each dimension to a pull coordinate index.
1840 This is only allowed if :mdp-value:`pull-coord1-type=external-potential` and
1841 :mdp:`pull-coord1-potential-provider` = ``awh`` for the concerned pull coordinate
1842 indices. Pull geometry 'direction-periodic' is not supported by AWH.
1844 .. mdp:: awh-potential
1846 .. mdp-value:: convolved
1848 The applied biasing potential is the convolution of the bias function and a
1849 set of harmonic umbrella potentials (see :mdp-value:`awh-potential=umbrella` below). This results
1850 in a smooth potential function and force. The resolution of the potential is set
1851 by the force constant of each umbrella, see :mdp:`awh1-dim1-force-constant`.
1853 .. mdp-value:: umbrella
1855 The potential bias is applied by controlling the position of an harmonic potential
1856 using Monte-Carlo sampling. The force constant is set with
1857 :mdp:`awh1-dim1-force-constant`. The umbrella location
1858 is sampled using Monte-Carlo every :mdp:`awh-nstsample` steps.
1859 There are no advantages to using an umbrella.
1860 This option is mainly for comparison and testing purposes.
1862 .. mdp:: awh-share-multisim
1866 AWH will not share biases across simulations started with
1867 :ref:`gmx mdrun` option ``-multidir``. The biases will be independent.
1871 With :ref:`gmx mdrun` and option ``-multidir`` the bias and PMF estimates
1872 for biases with :mdp:`awh1-share-group` >0 will be shared across simulations
1873 with the biases with the same :mdp:`awh1-share-group` value.
1874 The simulations should have the same AWH settings for sharing to make sense.
1875 :ref:`gmx mdrun` will check whether the simulations are technically
1876 compatible for sharing, but the user should check that bias sharing
1877 physically makes sense.
1881 (-1) Random seed for Monte-Carlo sampling the umbrella position,
1882 where -1 indicates to generate a seed. Only used with
1883 :mdp-value:`awh-potential=umbrella`.
1888 Number of steps between printing AWH data to the energy file, should be
1889 a multiple of :mdp:`nstenergy`.
1891 .. mdp:: awh-nstsample
1894 Number of steps between sampling of the coordinate value. This sampling
1895 is the basis for updating the bias and estimating the PMF and other AWH observables.
1897 .. mdp:: awh-nsamples-update
1900 The number of coordinate samples used for each AWH update.
1901 The update interval in steps is :mdp:`awh-nstsample` times this value.
1906 The number of biases, each acting on its own coordinate.
1907 The following options should be specified
1908 for each bias although below only the options for bias number 1 is shown. Options for
1909 other bias indices are obtained by replacing '1' by the bias index.
1911 .. mdp:: awh1-error-init
1913 (10.0) [kJ mol\ :sup:`-1`]
1914 Estimated initial average error of the PMF for this bias. This value together with the
1915 given diffusion constant(s) :mdp:`awh1-dim1-diffusion` determine the initial biasing rate.
1916 The error is obviously not known *a priori*. Only a rough estimate of :mdp:`awh1-error-init`
1918 As a general guideline, leave :mdp:`awh1-error-init` to its default value when starting a new
1919 simulation. On the other hand, when there is *a priori* knowledge of the PMF (e.g. when
1920 an initial PMF estimate is provided, see the :mdp:`awh1-user-data` option)
1921 then :mdp:`awh1-error-init` should reflect that knowledge.
1923 .. mdp:: awh1-growth
1925 .. mdp-value:: exp-linear
1927 Each bias keeps a reference weight histogram for the coordinate samples.
1928 Its size sets the magnitude of the bias function and free energy estimate updates
1929 (few samples corresponds to large updates and vice versa).
1930 Thus, its growth rate sets the maximum convergence rate.
1931 By default, there is an initial stage in which the histogram grows close to exponentially (but slower than the sampling rate).
1932 In the final stage that follows, the growth rate is linear and equal to the sampling rate (set by :mdp:`awh-nstsample`).
1933 The initial stage is typically necessary for efficient convergence when starting a new simulation where
1934 high free energy barriers have not yet been flattened by the bias.
1936 .. mdp-value:: linear
1938 As :mdp-value:`awh1-growth=exp-linear` but skip the initial stage. This may be useful if there is *a priori*
1939 knowledge (see :mdp:`awh1-error-init`) which eliminates the need for an initial stage. This is also
1940 the setting compatible with :mdp-value:`awh1-target=local-boltzmann`.
1942 .. mdp:: awh1-equilibrate-histogram
1946 Do not equilibrate histogram.
1950 Before entering the initial stage (see :mdp-value:`awh1-growth=exp-linear`), make sure the
1951 histogram of sampled weights is following the target distribution closely enough (specifically,
1952 at least 80% of the target region needs to have a local relative error of less than 20%). This
1953 option would typically only be used when :mdp:`awh1-share-group` > 0
1954 and the initial configurations poorly represent the target
1957 .. mdp:: awh1-target
1959 .. mdp-value:: constant
1961 The bias is tuned towards a constant (uniform) coordinate distribution
1962 in the defined sampling interval (defined by [:mdp:`awh1-dim1-start`, :mdp:`awh1-dim1-end`]).
1964 .. mdp-value:: cutoff
1966 Similar to :mdp-value:`awh1-target=constant`, but the target
1967 distribution is proportional to 1/(1 + exp(F - :mdp-value:`awh1-target=cutoff`)),
1968 where F is the free energy relative to the estimated global minimum.
1969 This provides a smooth switch of a flat target distribution in
1970 regions with free energy lower than the cut-off to a Boltzmann
1971 distribution in regions with free energy higher than the cut-off.
1973 .. mdp-value:: boltzmann
1975 The target distribution is a Boltzmann distribtution with a scaled beta (inverse temperature)
1976 factor given by :mdp:`awh1-target-beta-scaling`. *E.g.*, a value of 0.1
1977 would give the same coordinate distribution as sampling with a simulation temperature
1980 .. mdp-value:: local-boltzmann
1982 Same target distribution and use of :mdp:`awh1-target-beta-scaling`
1983 but the convergence towards the target distribution is inherently local *i.e.*, the rate of
1984 change of the bias only depends on the local sampling. This local convergence property is
1985 only compatible with :mdp-value:`awh1-growth=linear`, since for
1986 :mdp-value:`awh1-growth=exp-linear` histograms are globally rescaled in the initial stage.
1988 .. mdp:: awh1-target-beta-scaling
1991 For :mdp-value:`awh1-target=boltzmann` and :mdp-value:`awh1-target=local-boltzmann`
1992 it is the unitless beta scaling factor taking values in (0,1).
1994 .. mdp:: awh1-target-cutoff
1996 (0) [kJ mol\ :sup:`-1`]
1997 For :mdp-value:`awh1-target=cutoff` this is the cutoff, should be > 0.
1999 .. mdp:: awh1-user-data
2003 Initialize the PMF and target distribution with default values.
2007 Initialize the PMF and target distribution with user provided data. For :mdp:`awh-nbias` = 1,
2008 :ref:`gmx mdrun` will expect a file ``awhinit.xvg`` to be present in the run directory.
2009 For multiple biases, :ref:`gmx mdrun` expects files ``awhinit1.xvg``, ``awhinit2.xvg``, etc.
2010 The file name can be changed with the ``-awh`` option.
2011 The first :mdp:`awh1-ndim` columns of
2012 each input file should contain the coordinate values, such that each row defines a point in
2013 coordinate space. Column :mdp:`awh1-ndim` + 1 should contain the PMF value for each point.
2014 The target distribution column can either follow the PMF (column :mdp:`awh1-ndim` + 2) or
2015 be in the same column as written by :ref:`gmx awh`.
2017 .. mdp:: awh1-share-group
2021 Do not share the bias.
2023 .. mdp-value:: positive
2025 Share the bias and PMF estimates within and/or between simulations.
2026 Within a simulation, the bias will be shared between biases that have the
2027 same :mdp:`awh1-share-group` index (note that the current code does not support this).
2028 With :mdp-value:`awh-share-multisim=yes` and
2029 :ref:`gmx mdrun` option ``-multidir`` the bias will also be shared across simulations.
2030 Sharing may increase convergence initially, although the starting configurations
2031 can be critical, especially when sharing between many biases.
2032 Currently, positive group values should start at 1 and increase
2033 by 1 for each subsequent bias that is shared.
2038 Number of dimensions of the coordinate, each dimension maps to 1 pull coordinate.
2039 The following options should be specified for each such dimension. Below only
2040 the options for dimension number 1 is shown. Options for other dimension indices are
2041 obtained by replacing '1' by the dimension index.
2043 .. mdp:: awh1-dim1-coord-provider
2047 The module providing the reaction coordinate for this dimension.
2048 Currently AWH can only act on pull coordinates.
2050 .. mdp:: awh1-dim1-coord-index
2053 Index of the pull coordinate defining this coordinate dimension.
2055 .. mdp:: awh1-dim1-force-constant
2057 (0) [kJ mol\ :sup:`-1` nm\ :sup:`-2`] or [kJ mol\ :sup:`-1` rad\ :sup:`-2`]
2058 Force constant for the (convolved) umbrella potential(s) along this
2059 coordinate dimension.
2061 .. mdp:: awh1-dim1-start
2064 Start value of the sampling interval along this dimension. The range of allowed
2065 values depends on the relevant pull geometry (see :mdp:`pull-coord1-geometry`).
2066 For dihedral geometries :mdp:`awh1-dim1-start` greater than :mdp:`awh1-dim1-end`
2067 is allowed. The interval will then wrap around from +period/2 to -period/2.
2068 For the direction geometry, the dimension is made periodic when
2069 the direction is along a box vector and covers more than 95%
2070 of the box length. Note that one should not apply pressure coupling
2071 along a periodic dimension.
2073 .. mdp:: awh1-dim1-end
2076 End value defining the sampling interval together with :mdp:`awh1-dim1-start`.
2078 .. mdp:: awh1-dim1-diffusion
2080 (10\ :sup:`-5`) [nm\ :sup:`2`/ps] or [rad\ :sup:`2`/ps]
2081 Estimated diffusion constant for this coordinate dimension determining the initial
2082 biasing rate. This needs only be a rough estimate and should not critically
2083 affect the results unless it is set to something very low, leading to slow convergence,
2084 or very high, forcing the system far from equilibrium. Not setting this value
2085 explicitly generates a warning.
2087 .. mdp:: awh1-dim1-cover-diameter
2090 Diameter that needs to be sampled by a single simulation around a coordinate value
2091 before the point is considered covered in the initial stage (see :mdp-value:`awh1-growth=exp-linear`).
2092 A value > 0 ensures that for each covering there is a continuous transition of this diameter
2093 across each coordinate value.
2094 This is trivially true for independent simulations but not for for multiple bias-sharing simulations
2095 (:mdp:`awh1-share-group`>0).
2096 For a diameter = 0, covering occurs as soon as the simulations have sampled the whole interval, which
2097 for many sharing simulations does not guarantee transitions across free energy barriers.
2098 On the other hand, when the diameter >= the sampling interval length, covering occurs when a single simulation
2099 has independently sampled the whole interval.
2104 These :ref:`mdp` parameters can be used enforce the rotation of a group of atoms,
2105 e.g. a protein subunit. The `reference manual`_ describes in detail 13 different potentials
2106 that can be used to achieve such a rotation.
2112 No enforced rotation will be applied. All enforced rotation options will
2113 be ignored (and if present in the :ref:`mdp` file, they unfortunately
2118 Apply the rotation potential specified by :mdp:`rot-type0` to the group of atoms given
2119 under the :mdp:`rot-group0` option.
2121 .. mdp:: rot-ngroups
2124 Number of rotation groups.
2128 Name of rotation group 0 in the index file.
2133 Type of rotation potential that is applied to rotation group 0. Can be of of the following:
2134 ``iso``, ``iso-pf``, ``pm``, ``pm-pf``, ``rm``, ``rm-pf``, ``rm2``, ``rm2-pf``,
2135 ``flex``, ``flex-t``, ``flex2``, or ``flex2-t``.
2140 Use mass weighted rotation group positions.
2145 Rotation vector, will get normalized.
2150 Pivot point for the potentials ``iso``, ``pm``, ``rm``, and ``rm2``.
2154 (0) [degree ps\ :sup:`-1`]
2155 Reference rotation rate of group 0.
2159 (0) [kJ mol\ :sup:`-1` nm\ :sup:`-2`]
2160 Force constant for group 0.
2162 .. mdp:: rot-slab-dist0
2165 Slab distance, if a flexible axis rotation type was chosen.
2167 .. mdp:: rot-min-gauss0
2170 Minimum value (cutoff) of Gaussian function for the force to be evaluated
2171 (for the flexible axis potentials).
2175 (0.0001) [nm\ :sup:`2`]
2176 Value of additive constant epsilon for ``rm2*`` and ``flex2*`` potentials.
2178 .. mdp:: rot-fit-method0
2181 Fitting method when determining the actual angle of a rotation group
2182 (can be one of ``rmsd``, ``norm``, or ``potential``).
2184 .. mdp:: rot-potfit-nsteps0
2187 For fit type ``potential``, the number of angular positions around the reference angle for which the
2188 rotation potential is evaluated.
2190 .. mdp:: rot-potfit-step0
2193 For fit type ``potential``, the distance in degrees between two angular positions.
2195 .. mdp:: rot-nstrout
2198 Output frequency (in steps) for the angle of the rotation group, as well as for the torque
2199 and the rotation potential energy.
2201 .. mdp:: rot-nstsout
2204 Output frequency for per-slab data of the flexible axis potentials, i.e. angles, torques and slab centers.
2214 ignore distance restraint information in topology file
2216 .. mdp-value:: simple
2218 simple (per-molecule) distance restraints.
2220 .. mdp-value:: ensemble
2222 distance restraints over an ensemble of molecules in one
2223 simulation box. Normally, one would perform ensemble averaging
2224 over multiple simulations, using ``mdrun
2225 -multidir``. The environment
2226 variable ``GMX_DISRE_ENSEMBLE_SIZE`` sets the number of systems
2227 within each ensemble (usually equal to the number of directories
2228 supplied to ``mdrun -multidir``).
2230 .. mdp:: disre-weighting
2232 .. mdp-value:: equal
2234 divide the restraint force equally over all atom pairs in the
2237 .. mdp-value:: conservative
2239 the forces are the derivative of the restraint potential, this
2240 results in an weighting of the atom pairs to the reciprocal
2241 seventh power of the displacement. The forces are conservative
2242 when :mdp:`disre-tau` is zero.
2244 .. mdp:: disre-mixed
2248 the violation used in the calculation of the restraint force is
2249 the time-averaged violation
2253 the violation used in the calculation of the restraint force is
2254 the square root of the product of the time-averaged violation
2255 and the instantaneous violation
2259 (1000) [kJ mol\ :sup:`-1` nm\ :sup:`-2`]
2260 force constant for distance restraints, which is multiplied by a
2261 (possibly) different factor for each restraint given in the `fac`
2262 column of the interaction in the topology file.
2267 time constant for distance restraints running average. A value of
2268 zero turns off time averaging.
2270 .. mdp:: nstdisreout
2273 period between steps when the running time-averaged and
2274 instantaneous distances of all atom pairs involved in restraints
2275 are written to the energy file (can make the energy file very
2282 ignore orientation restraint information in topology file
2286 use orientation restraints, ensemble averaging can be performed
2287 with ``mdrun -multidir``
2291 (0) [kJ mol\ :sup:`-1`]
2292 force constant for orientation restraints, which is multiplied by a
2293 (possibly) different weight factor for each restraint, can be set
2294 to zero to obtain the orientations from a free simulation
2299 time constant for orientation restraints running average. A value
2300 of zero turns off time averaging.
2302 .. mdp:: orire-fitgrp
2304 fit group for orientation restraining. This group of atoms is used
2305 to determine the rotation **R** of the system with respect to the
2306 reference orientation. The reference orientation is the starting
2307 conformation of the first subsystem. For a protein, backbone is a
2310 .. mdp:: nstorireout
2313 period between steps when the running time-averaged and
2314 instantaneous orientations for all restraints, and the molecular
2315 order tensor are written to the energy file (can make the energy
2319 Free energy calculations
2320 ^^^^^^^^^^^^^^^^^^^^^^^^
2322 .. mdp:: free-energy
2326 Only use topology A.
2330 Interpolate between topology A (lambda=0) to topology B
2331 (lambda=1) and write the derivative of the Hamiltonian with
2332 respect to lambda (as specified with :mdp:`dhdl-derivatives`),
2333 or the Hamiltonian differences with respect to other lambda
2334 values (as specified with foreign lambda) to the energy file
2335 and/or to ``dhdl.xvg``, where they can be processed by, for
2336 example :ref:`gmx bar`. The potentials, bond-lengths and angles
2337 are interpolated linearly as described in the manual. When
2338 :mdp:`sc-alpha` is larger than zero, soft-core potentials are
2339 used for the LJ and Coulomb interactions.
2343 Turns on expanded ensemble simulation, where the alchemical state
2344 becomes a dynamic variable, allowing jumping between different
2345 Hamiltonians. See the expanded ensemble options for controlling how
2346 expanded ensemble simulations are performed. The different
2347 Hamiltonians used in expanded ensemble simulations are defined by
2348 the other free energy options.
2350 .. mdp:: init-lambda
2353 starting value for lambda (float). Generally, this should only be
2354 used with slow growth (*i.e.* nonzero :mdp:`delta-lambda`). In
2355 other cases, :mdp:`init-lambda-state` should be specified
2356 instead. Must be greater than or equal to 0.
2358 .. mdp:: delta-lambda
2361 increment per time step for lambda
2363 .. mdp:: init-lambda-state
2366 starting value for the lambda state (integer). Specifies which
2367 columm of the lambda vector (:mdp:`coul-lambdas`,
2368 :mdp:`vdw-lambdas`, :mdp:`bonded-lambdas`,
2369 :mdp:`restraint-lambdas`, :mdp:`mass-lambdas`,
2370 :mdp:`temperature-lambdas`, :mdp:`fep-lambdas`) should be
2371 used. This is a zero-based index: :mdp:`init-lambda-state` 0 means
2372 the first column, and so on.
2374 .. mdp:: fep-lambdas
2377 Zero, one or more lambda values for which Delta H values will be
2378 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2379 steps. Values must be between 0 and 1. Free energy differences
2380 between different lambda values can then be determined with
2381 :ref:`gmx bar`. :mdp:`fep-lambdas` is different from the
2382 other -lambdas keywords because all components of the lambda vector
2383 that are not specified will use :mdp:`fep-lambdas` (including
2384 :mdp:`restraint-lambdas` and therefore the pull code restraints).
2386 .. mdp:: coul-lambdas
2389 Zero, one or more lambda values for which Delta H values will be
2390 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2391 steps. Values must be between 0 and 1. Only the electrostatic
2392 interactions are controlled with this component of the lambda
2393 vector (and only if the lambda=0 and lambda=1 states have differing
2394 electrostatic interactions).
2396 .. mdp:: vdw-lambdas
2399 Zero, one or more lambda values for which Delta H values will be
2400 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2401 steps. Values must be between 0 and 1. Only the van der Waals
2402 interactions are controlled with this component of the lambda
2405 .. mdp:: bonded-lambdas
2408 Zero, one or more lambda values for which Delta H values will be
2409 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2410 steps. Values must be between 0 and 1. Only the bonded interactions
2411 are controlled with this component of the lambda vector.
2413 .. mdp:: restraint-lambdas
2416 Zero, one or more lambda values for which Delta H values will be
2417 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2418 steps. Values must be between 0 and 1. Only the restraint
2419 interactions: dihedral restraints, and the pull code restraints are
2420 controlled with this component of the lambda vector.
2422 .. mdp:: mass-lambdas
2425 Zero, one or more lambda values for which Delta H values will be
2426 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2427 steps. Values must be between 0 and 1. Only the particle masses are
2428 controlled with this component of the lambda vector.
2430 .. mdp:: temperature-lambdas
2433 Zero, one or more lambda values for which Delta H values will be
2434 determined and written to dhdl.xvg every :mdp:`nstdhdl`
2435 steps. Values must be between 0 and 1. Only the temperatures
2436 controlled with this component of the lambda vector. Note that
2437 these lambdas should not be used for replica exchange, only for
2438 simulated tempering.
2440 .. mdp:: calc-lambda-neighbors
2443 Controls the number of lambda values for which Delta H values will
2444 be calculated and written out, if :mdp:`init-lambda-state` has
2445 been set. A positive value will limit the number of lambda points
2446 calculated to only the nth neighbors of :mdp:`init-lambda-state`:
2447 for example, if :mdp:`init-lambda-state` is 5 and this parameter
2448 has a value of 2, energies for lambda points 3-7 will be calculated
2449 and writen out. A value of -1 means all lambda points will be
2450 written out. For normal BAR such as with :ref:`gmx bar`, a value of
2451 1 is sufficient, while for MBAR -1 should be used.
2456 the soft-core alpha parameter, a value of 0 results in linear
2457 interpolation of the LJ and Coulomb interactions
2462 the power of the radial term in the soft-core equation. Possible
2463 values are 6 and 48. 6 is more standard, and is the default. When
2464 48 is used, then sc-alpha should generally be much lower (between
2470 Whether to apply the soft-core free energy interaction
2471 transformation to the Columbic interaction of a molecule. Default
2472 is no, as it is generally more efficient to turn off the Coulomic
2473 interactions linearly before turning off the van der Waals
2474 interactions. Note that it is only taken into account when lambda
2475 states are used, not with :mdp:`couple-lambda0` /
2476 :mdp:`couple-lambda1`, and you can still turn off soft-core
2477 interactions by setting :mdp:`sc-alpha` to 0.
2482 the power for lambda in the soft-core function, only the values 1
2488 the soft-core sigma for particles which have a C6 or C12 parameter
2489 equal to zero or a sigma smaller than :mdp:`sc-sigma`
2491 .. mdp:: couple-moltype
2493 Here one can supply a molecule type (as defined in the topology)
2494 for calculating solvation or coupling free energies. There is a
2495 special option ``system`` that couples all molecule types in the
2496 system. This can be useful for equilibrating a system starting from
2497 (nearly) random coordinates. :mdp:`free-energy` has to be turned
2498 on. The Van der Waals interactions and/or charges in this molecule
2499 type can be turned on or off between lambda=0 and lambda=1,
2500 depending on the settings of :mdp:`couple-lambda0` and
2501 :mdp:`couple-lambda1`. If you want to decouple one of several
2502 copies of a molecule, you need to copy and rename the molecule
2503 definition in the topology.
2505 .. mdp:: couple-lambda0
2507 .. mdp-value:: vdw-q
2509 all interactions are on at lambda=0
2513 the charges are zero (no Coulomb interactions) at lambda=0
2517 the Van der Waals interactions are turned at lambda=0; soft-core
2518 interactions will be required to avoid singularities
2522 the Van der Waals interactions are turned off and the charges
2523 are zero at lambda=0; soft-core interactions will be required to
2524 avoid singularities.
2526 .. mdp:: couple-lambda1
2528 analogous to :mdp:`couple-lambda1`, but for lambda=1
2530 .. mdp:: couple-intramol
2534 All intra-molecular non-bonded interactions for moleculetype
2535 :mdp:`couple-moltype` are replaced by exclusions and explicit
2536 pair interactions. In this manner the decoupled state of the
2537 molecule corresponds to the proper vacuum state without
2538 periodicity effects.
2542 The intra-molecular Van der Waals and Coulomb interactions are
2543 also turned on/off. This can be useful for partitioning
2544 free-energies of relatively large molecules, where the
2545 intra-molecular non-bonded interactions might lead to
2546 kinetically trapped vacuum conformations. The 1-4 pair
2547 interactions are not turned off.
2552 the frequency for writing dH/dlambda and possibly Delta H to
2553 dhdl.xvg, 0 means no ouput, should be a multiple of
2554 :mdp:`nstcalcenergy`.
2556 .. mdp:: dhdl-derivatives
2560 If yes (the default), the derivatives of the Hamiltonian with
2561 respect to lambda at each :mdp:`nstdhdl` step are written
2562 out. These values are needed for interpolation of linear energy
2563 differences with :ref:`gmx bar` (although the same can also be
2564 achieved with the right foreign lambda setting, that may not be as
2565 flexible), or with thermodynamic integration
2567 .. mdp:: dhdl-print-energy
2571 Include either the total or the potential energy in the dhdl
2572 file. Options are 'no', 'potential', or 'total'. This information
2573 is needed for later free energy analysis if the states of interest
2574 are at different temperatures. If all states are at the same
2575 temperature, this information is not needed. 'potential' is useful
2576 in case one is using ``mdrun -rerun`` to generate the ``dhdl.xvg``
2577 file. When rerunning from an existing trajectory, the kinetic
2578 energy will often not be correct, and thus one must compute the
2579 residual free energy from the potential alone, with the kinetic
2580 energy component computed analytically.
2582 .. mdp:: separate-dhdl-file
2586 The free energy values that are calculated (as specified with
2587 the foreign lambda and :mdp:`dhdl-derivatives` settings) are
2588 written out to a separate file, with the default name
2589 ``dhdl.xvg``. This file can be used directly with :ref:`gmx
2594 The free energy values are written out to the energy output file
2595 (``ener.edr``, in accumulated blocks at every :mdp:`nstenergy`
2596 steps), where they can be extracted with :ref:`gmx energy` or
2597 used directly with :ref:`gmx bar`.
2599 .. mdp:: dh-hist-size
2602 If nonzero, specifies the size of the histogram into which the
2603 Delta H values (specified with foreign lambda) and the derivative
2604 dH/dl values are binned, and written to ener.edr. This can be used
2605 to save disk space while calculating free energy differences. One
2606 histogram gets written for each foreign lambda and two for the
2607 dH/dl, at every :mdp:`nstenergy` step. Be aware that incorrect
2608 histogram settings (too small size or too wide bins) can introduce
2609 errors. Do not use histograms unless you're certain you need it.
2611 .. mdp:: dh-hist-spacing
2614 Specifies the bin width of the histograms, in energy units. Used in
2615 conjunction with :mdp:`dh-hist-size`. This size limits the
2616 accuracy with which free energies can be calculated. Do not use
2617 histograms unless you're certain you need it.
2620 Expanded Ensemble calculations
2621 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2623 .. mdp:: nstexpanded
2625 The number of integration steps beween attempted moves changing the
2626 system Hamiltonian in expanded ensemble simulations. Must be a
2627 multiple of :mdp:`nstcalcenergy`, but can be greater or less than
2634 No Monte Carlo in state space is performed.
2636 .. mdp-value:: metropolis-transition
2638 Uses the Metropolis weights to update the expanded ensemble
2639 weight of each state. Min{1,exp(-(beta_new u_new - beta_old
2642 .. mdp-value:: barker-transition
2644 Uses the Barker transition critera to update the expanded
2645 ensemble weight of each state i, defined by exp(-beta_new
2646 u_new)/(exp(-beta_new u_new)+exp(-beta_old u_old))
2648 .. mdp-value:: wang-landau
2650 Uses the Wang-Landau algorithm (in state space, not energy
2651 space) to update the expanded ensemble weights.
2653 .. mdp-value:: min-variance
2655 Uses the minimum variance updating method of Escobedo et al. to
2656 update the expanded ensemble weights. Weights will not be the
2657 free energies, but will rather emphasize states that need more
2658 sampling to give even uncertainty.
2660 .. mdp:: lmc-mc-move
2664 No Monte Carlo in state space is performed.
2666 .. mdp-value:: metropolis-transition
2668 Randomly chooses a new state up or down, then uses the
2669 Metropolis critera to decide whether to accept or reject:
2670 Min{1,exp(-(beta_new u_new - beta_old u_old)}
2672 .. mdp-value:: barker-transition
2674 Randomly chooses a new state up or down, then uses the Barker
2675 transition critera to decide whether to accept or reject:
2676 exp(-beta_new u_new)/(exp(-beta_new u_new)+exp(-beta_old u_old))
2678 .. mdp-value:: gibbs
2680 Uses the conditional weights of the state given the coordinate
2681 (exp(-beta_i u_i) / sum_k exp(beta_i u_i) to decide which state
2684 .. mdp-value:: metropolized-gibbs
2686 Uses the conditional weights of the state given the coordinate
2687 (exp(-beta_i u_i) / sum_k exp(beta_i u_i) to decide which state
2688 to move to, EXCLUDING the current state, then uses a rejection
2689 step to ensure detailed balance. Always more efficient that
2690 Gibbs, though only marginally so in many situations, such as
2691 when only the nearest neighbors have decent phase space
2697 random seed to use for Monte Carlo moves in state space. When
2698 :mdp:`lmc-seed` is set to -1, a pseudo random seed is us
2700 .. mdp:: mc-temperature
2702 Temperature used for acceptance/rejection for Monte Carlo moves. If
2703 not specified, the temperature of the simulation specified in the
2704 first group of :mdp:`ref-t` is used.
2709 The cutoff for the histogram of state occupancies to be reset, and
2710 the free energy incrementor to be changed from delta to delta *
2711 :mdp:`wl-scale`. If we define the Nratio = (number of samples at
2712 each histogram) / (average number of samples at each
2713 histogram). :mdp:`wl-ratio` of 0.8 means that means that the
2714 histogram is only considered flat if all Nratio > 0.8 AND
2715 simultaneously all 1/Nratio > 0.8.
2720 Each time the histogram is considered flat, then the current value
2721 of the Wang-Landau incrementor for the free energies is multiplied
2722 by :mdp:`wl-scale`. Value must be between 0 and 1.
2724 .. mdp:: init-wl-delta
2727 The initial value of the Wang-Landau incrementor in kT. Some value
2728 near 1 kT is usually most efficient, though sometimes a value of
2729 2-3 in units of kT works better if the free energy differences are
2732 .. mdp:: wl-oneovert
2735 Set Wang-Landau incrementor to scale with 1/(simulation time) in
2736 the large sample limit. There is significant evidence that the
2737 standard Wang-Landau algorithms in state space presented here
2738 result in free energies getting 'burned in' to incorrect values
2739 that depend on the initial state. when :mdp:`wl-oneovert` is true,
2740 then when the incrementor becomes less than 1/N, where N is the
2741 mumber of samples collected (and thus proportional to the data
2742 collection time, hence '1 over t'), then the Wang-Lambda
2743 incrementor is set to 1/N, decreasing every step. Once this occurs,
2744 :mdp:`wl-ratio` is ignored, but the weights will still stop
2745 updating when the equilibration criteria set in
2746 :mdp:`lmc-weights-equil` is achieved.
2748 .. mdp:: lmc-repeats
2751 Controls the number of times that each Monte Carlo swap type is
2752 performed each iteration. In the limit of large numbers of Monte
2753 Carlo repeats, then all methods converge to Gibbs sampling. The
2754 value will generally not need to be different from 1.
2756 .. mdp:: lmc-gibbsdelta
2759 Limit Gibbs sampling to selected numbers of neighboring states. For
2760 Gibbs sampling, it is sometimes inefficient to perform Gibbs
2761 sampling over all of the states that are defined. A positive value
2762 of :mdp:`lmc-gibbsdelta` means that only states plus or minus
2763 :mdp:`lmc-gibbsdelta` are considered in exchanges up and down. A
2764 value of -1 means that all states are considered. For less than 100
2765 states, it is probably not that expensive to include all states.
2767 .. mdp:: lmc-forced-nstart
2770 Force initial state space sampling to generate weights. In order to
2771 come up with reasonable initial weights, this setting allows the
2772 simulation to drive from the initial to the final lambda state,
2773 with :mdp:`lmc-forced-nstart` steps at each state before moving on
2774 to the next lambda state. If :mdp:`lmc-forced-nstart` is
2775 sufficiently long (thousands of steps, perhaps), then the weights
2776 will be close to correct. However, in most cases, it is probably
2777 better to simply run the standard weight equilibration algorithms.
2779 .. mdp:: nst-transition-matrix
2782 Frequency of outputting the expanded ensemble transition matrix. A
2783 negative number means it will only be printed at the end of the
2786 .. mdp:: symmetrized-transition-matrix
2789 Whether to symmetrize the empirical transition matrix. In the
2790 infinite limit the matrix will be symmetric, but will diverge with
2791 statistical noise for short timescales. Forced symmetrization, by
2792 using the matrix T_sym = 1/2 (T + transpose(T)), removes problems
2793 like the existence of (small magnitude) negative eigenvalues.
2795 .. mdp:: mininum-var-min
2798 The min-variance strategy (option of :mdp:`lmc-stats` is only
2799 valid for larger number of samples, and can get stuck if too few
2800 samples are used at each state. :mdp:`mininum-var-min` is the
2801 minimum number of samples that each state that are allowed before
2802 the min-variance strategy is activated if selected.
2804 .. mdp:: init-lambda-weights
2806 The initial weights (free energies) used for the expanded ensemble
2807 states. Default is a vector of zero weights. format is similar to
2808 the lambda vector settings in :mdp:`fep-lambdas`, except the
2809 weights can be any floating point number. Units are kT. Its length
2810 must match the lambda vector lengths.
2812 .. mdp:: lmc-weights-equil
2816 Expanded ensemble weights continue to be updated throughout the
2821 The input expanded ensemble weights are treated as equilibrated,
2822 and are not updated throughout the simulation.
2824 .. mdp-value:: wl-delta
2826 Expanded ensemble weight updating is stopped when the
2827 Wang-Landau incrementor falls below this value.
2829 .. mdp-value:: number-all-lambda
2831 Expanded ensemble weight updating is stopped when the number of
2832 samples at all of the lambda states is greater than this value.
2834 .. mdp-value:: number-steps
2836 Expanded ensemble weight updating is stopped when the number of
2837 steps is greater than the level specified by this value.
2839 .. mdp-value:: number-samples
2841 Expanded ensemble weight updating is stopped when the number of
2842 total samples across all lambda states is greater than the level
2843 specified by this value.
2845 .. mdp-value:: count-ratio
2847 Expanded ensemble weight updating is stopped when the ratio of
2848 samples at the least sampled lambda state and most sampled
2849 lambda state greater than this value.
2851 .. mdp:: simulated-tempering
2854 Turn simulated tempering on or off. Simulated tempering is
2855 implemented as expanded ensemble sampling with different
2856 temperatures instead of different Hamiltonians.
2858 .. mdp:: sim-temp-low
2861 Low temperature for simulated tempering.
2863 .. mdp:: sim-temp-high
2866 High temperature for simulated tempering.
2868 .. mdp:: simulated-tempering-scaling
2870 Controls the way that the temperatures at intermediate lambdas are
2871 calculated from the :mdp:`temperature-lambdas` part of the lambda
2874 .. mdp-value:: linear
2876 Linearly interpolates the temperatures using the values of
2877 :mdp:`temperature-lambdas`, *i.e.* if :mdp:`sim-temp-low`
2878 =300, :mdp:`sim-temp-high` =400, then lambda=0.5 correspond to
2879 a temperature of 350. A nonlinear set of temperatures can always
2880 be implemented with uneven spacing in lambda.
2882 .. mdp-value:: geometric
2884 Interpolates temperatures geometrically between
2885 :mdp:`sim-temp-low` and :mdp:`sim-temp-high`. The i:th state
2886 has temperature :mdp:`sim-temp-low` * (:mdp:`sim-temp-high` /
2887 :mdp:`sim-temp-low`) raised to the power of
2888 (i/(ntemps-1)). This should give roughly equal exchange for
2889 constant heat capacity, though of course things simulations that
2890 involve protein folding have very high heat capacity peaks.
2892 .. mdp-value:: exponential
2894 Interpolates temperatures exponentially between
2895 :mdp:`sim-temp-low` and :mdp:`sim-temp-high`. The i:th state
2896 has temperature :mdp:`sim-temp-low` + (:mdp:`sim-temp-high` -
2897 :mdp:`sim-temp-low`)*((exp(:mdp:`temperature-lambdas`
2898 (i))-1)/(exp(1.0)-i)).
2906 groups for constant acceleration (*e.g.* ``Protein Sol``) all atoms
2907 in groups Protein and Sol will experience constant acceleration as
2908 specified in the :mdp:`accelerate` line
2912 (0) [nm ps\ :sup:`-2`]
2913 acceleration for :mdp:`acc-grps`; x, y and z for each group
2914 (*e.g.* ``0.1 0.0 0.0 -0.1 0.0 0.0`` means that first group has
2915 constant acceleration of 0.1 nm ps\ :sup:`-2` in X direction, second group
2920 Groups that are to be frozen (*i.e.* their X, Y, and/or Z position
2921 will not be updated; *e.g.* ``Lipid SOL``). :mdp:`freezedim`
2922 specifies for which dimension(s) the freezing applies. To avoid
2923 spurious contributions to the virial and pressure due to large
2924 forces between completely frozen atoms you need to use energy group
2925 exclusions, this also saves computing time. Note that coordinates
2926 of frozen atoms are not scaled by pressure-coupling algorithms.
2930 dimensions for which groups in :mdp:`freezegrps` should be frozen,
2931 specify `Y` or `N` for X, Y and Z and for each group (*e.g.* ``Y Y
2932 N N N N`` means that particles in the first group can move only in
2933 Z direction. The particles in the second group can move in any
2936 .. mdp:: cos-acceleration
2938 (0) [nm ps\ :sup:`-2`]
2939 the amplitude of the acceleration profile for calculating the
2940 viscosity. The acceleration is in the X-direction and the magnitude
2941 is :mdp:`cos-acceleration` cos(2 pi z/boxheight). Two terms are
2942 added to the energy file: the amplitude of the velocity profile and
2947 (0 0 0 0 0 0) [nm ps\ :sup:`-1`]
2948 The velocities of deformation for the box elements: a(x) b(y) c(z)
2949 b(x) c(x) c(y). Each step the box elements for which :mdp:`deform`
2950 is non-zero are calculated as: box(ts)+(t-ts)*deform, off-diagonal
2951 elements are corrected for periodicity. The coordinates are
2952 transformed accordingly. Frozen degrees of freedom are (purposely)
2953 also transformed. The time ts is set to t at the first step and at
2954 steps at which x and v are written to trajectory to ensure exact
2955 restarts. Deformation can be used together with semiisotropic or
2956 anisotropic pressure coupling when the appropriate
2957 compressibilities are set to zero. The diagonal elements can be
2958 used to strain a solid. The off-diagonal elements can be used to
2959 shear a solid or a liquid.
2965 .. mdp:: electric-field-x
2966 .. mdp:: electric-field-y
2967 .. mdp:: electric-field-z
2969 Here you can specify an electric field that optionally can be
2970 alternating and pulsed. The general expression for the field
2971 has the form of a gaussian laser pulse:
2973 .. math:: E(t) = E_0 \exp\left[-\frac{(t-t_0)^2}{2\sigma^2}\right]\cos\left[\omega (t-t_0)\right]
2975 For example, the four parameters for direction x are set in the
2976 fields of :mdp:`electric-field-x` (and similar for ``electric-field-y``
2977 and ``electric-field-z``) like
2979 ``electric-field-x = E0 omega t0 sigma``
2981 with units (respectively) V nm\ :sup:`-1`, ps\ :sup:`-1`, ps, ps.
2983 In the special case that ``sigma = 0``, the exponential term is omitted
2984 and only the cosine term is used. If also ``omega = 0`` a static
2985 electric field is applied.
2987 Read more at :ref:`electric fields` and in ref. \ :ref:`146 <refCaleman2008a>`.
2990 Mixed quantum/classical molecular dynamics
2991 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3001 Do a QM/MM simulation. Several groups can be described at
3002 different QM levels separately. These are specified in the
3003 :mdp:`QMMM-grps` field separated by spaces. The level of *ab
3004 initio* theory at which the groups are described is specified by
3005 :mdp:`QMmethod` and :mdp:`QMbasis` Fields. Describing the
3006 groups at different levels of theory is only possible with the
3007 ONIOM QM/MM scheme, specified by :mdp:`QMMMscheme`.
3011 groups to be descibed at the QM level (works also in case of MiMiC QM/MM)
3015 .. mdp-value:: normal
3017 normal QM/MM. There can only be one :mdp:`QMMM-grps` that is
3018 modelled at the :mdp:`QMmethod` and :mdp:`QMbasis` level of
3019 *ab initio* theory. The rest of the system is described at the
3020 MM level. The QM and MM subsystems interact as follows: MM point
3021 charges are included in the QM one-electron hamiltonian and all
3022 Lennard-Jones interactions are described at the MM level.
3024 .. mdp-value:: ONIOM
3026 The interaction between the subsystem is described using the
3027 ONIOM method by Morokuma and co-workers. There can be more than
3028 one :mdp:`QMMM-grps` each modeled at a different level of QM
3029 theory (:mdp:`QMmethod` and :mdp:`QMbasis`).
3034 Method used to compute the energy and gradients on the QM
3035 atoms. Available methods are AM1, PM3, RHF, UHF, DFT, B3LYP, MP2,
3036 CASSCF, and MMVB. For CASSCF, the number of electrons and orbitals
3037 included in the active space is specified by :mdp:`CASelectrons`
3038 and :mdp:`CASorbitals`.
3043 Basis set used to expand the electronic wavefuntion. Only Gaussian
3044 basis sets are currently available, *i.e.* ``STO-3G, 3-21G, 3-21G*,
3045 3-21+G*, 6-21G, 6-31G, 6-31G*, 6-31+G*,`` and ``6-311G``.
3050 The total charge in `e` of the :mdp:`QMMM-grps`. In case there are
3051 more than one :mdp:`QMMM-grps`, the total charge of each ONIOM
3052 layer needs to be specified separately.
3057 The multiplicity of the :mdp:`QMMM-grps`. In case there are more
3058 than one :mdp:`QMMM-grps`, the multiplicity of each ONIOM layer
3059 needs to be specified separately.
3061 .. mdp:: CASorbitals
3064 The number of orbitals to be included in the active space when
3065 doing a CASSCF computation.
3067 .. mdp:: CASelectrons
3070 The number of electrons to be included in the active space when
3071 doing a CASSCF computation.
3077 No surface hopping. The system is always in the electronic
3082 Do a QM/MM MD simulation on the excited state-potential energy
3083 surface and enforce a *diabatic* hop to the ground-state when
3084 the system hits the conical intersection hyperline in the course
3085 the simulation. This option only works in combination with the
3089 Computational Electrophysiology
3090 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3091 Use these options to switch on and control ion/water position exchanges in "Computational
3092 Electrophysiology" simulation setups. (See the `reference manual`_ for details).
3098 Do not enable ion/water position exchanges.
3100 .. mdp-value:: X ; Y ; Z
3102 Allow for ion/water position exchanges along the chosen direction.
3103 In a typical setup with the membranes parallel to the x-y plane,
3104 ion/water pairs need to be exchanged in Z direction to sustain the
3105 requested ion concentrations in the compartments.
3107 .. mdp:: swap-frequency
3109 (1) The swap attempt frequency, i.e. every how many time steps the ion counts
3110 per compartment are determined and exchanges made if necessary.
3111 Normally it is not necessary to check at every time step.
3112 For typical Computational Electrophysiology setups, a value of about 100 is
3113 sufficient and yields a negligible performance impact.
3115 .. mdp:: split-group0
3117 Name of the index group of the membrane-embedded part of channel #0.
3118 The center of mass of these atoms defines one of the compartment boundaries
3119 and should be chosen such that it is near the center of the membrane.
3121 .. mdp:: split-group1
3123 Channel #1 defines the position of the other compartment boundary.
3125 .. mdp:: massw-split0
3127 (no) Defines whether or not mass-weighting is used to calculate the split group center.
3131 Use the geometrical center.
3135 Use the center of mass.
3137 .. mdp:: massw-split1
3139 (no) As above, but for split-group #1.
3141 .. mdp:: solvent-group
3143 Name of the index group of solvent molecules.
3145 .. mdp:: coupl-steps
3147 (10) Average the number of ions per compartment over these many swap attempt steps.
3148 This can be used to prevent that ions near a compartment boundary
3149 (diffusing through a channel, e.g.) lead to unwanted back and forth swaps.
3153 (1) The number of different ion types to be controlled. These are during the
3154 simulation exchanged with solvent molecules to reach the desired reference numbers.
3156 .. mdp:: iontype0-name
3158 Name of the first ion type.
3160 .. mdp:: iontype0-in-A
3162 (-1) Requested (=reference) number of ions of type 0 in compartment A.
3163 The default value of -1 means: use the number of ions as found in time step 0
3166 .. mdp:: iontype0-in-B
3168 (-1) Reference number of ions of type 0 for compartment B.
3170 .. mdp:: bulk-offsetA
3172 (0.0) Offset of the first swap layer from the compartment A midplane.
3173 By default (i.e. bulk offset = 0.0), ion/water exchanges happen between layers
3174 at maximum distance (= bulk concentration) to the split group layers. However,
3175 an offset b (-1.0 < b < +1.0) can be specified to offset the bulk layer from the middle at 0.0
3176 towards one of the compartment-partitioning layers (at +/- 1.0).
3178 .. mdp:: bulk-offsetB
3180 (0.0) Offset of the other swap layer from the compartment B midplane.
3185 (\1) Only swap ions if threshold difference to requested count is reached.
3189 (2.0) [nm] Radius of the split cylinder #0.
3190 Two split cylinders (mimicking the channel pores) can optionally be defined
3191 relative to the center of the split group. With the help of these cylinders
3192 it can be counted which ions have passed which channel. The split cylinder
3193 definition has no impact on whether or not ion/water swaps are done.
3197 (1.0) [nm] Upper extension of the split cylinder #0.
3201 (1.0) [nm] Lower extension of the split cylinder #0.
3205 (2.0) [nm] Radius of the split cylinder #1.
3209 (1.0) [nm] Upper extension of the split cylinder #1.
3213 (1.0) [nm] Lower extension of the split cylinder #1.
3216 User defined thingies
3217 ^^^^^^^^^^^^^^^^^^^^^
3221 .. mdp:: userint1 (0)
3222 .. mdp:: userint2 (0)
3223 .. mdp:: userint3 (0)
3224 .. mdp:: userint4 (0)
3225 .. mdp:: userreal1 (0)
3226 .. mdp:: userreal2 (0)
3227 .. mdp:: userreal3 (0)
3228 .. mdp:: userreal4 (0)
3230 These you can use if you modify code. You can pass integers and
3231 reals and groups to your subroutine. Check the inputrec definition
3232 in ``src/gromacs/mdtypes/inputrec.h``
3237 These features have been removed from |Gromacs|, but so that old
3238 :ref:`mdp` and :ref:`tpr` files cannot be mistakenly misused, we still
3239 parse this option. :ref:`gmx grompp` and :ref:`gmx mdrun` will issue a
3240 fatal error if this is set.
3246 .. mdp:: implicit-solvent
3250 .. _reference manual: gmx-manual-parent-dir_