This is flucq, produced by makeinfo version 4.0 from flucq.texi.CHARMM Element doc/flucq.doc $Revision: 1.2 $#File: flucq, Node: Top, Next: Syntax, Up: (chmdoc/commands.doc) Combined QM/MM Fluctuating Charge Potential for CHARMM Ben Webb, [email protected], and Paul Lyne The fluctuating charge potential (FlucQ or FQ) is based on the methoddeveloped by Rick, Stuart and Berne (Rick et. al., J. Chem. Phys. 101(7) 1994 p6141) for molecular dynamics, and extended for hybrid QM/MMsimulations (Bryce et. al., Chem. Phys. Lett. 279 1997, p367). It isdesigned primarily for computationally efficient (approx. 10% overhead)modelling of solvent polarisation in hybrid QM/MM systems, and as suchis implemented for QUANTUM, CADPAC and GAMESS codes, although thecurrent implementation is easily extensible to any atom type and bond.* Menu:* Syntax:: Syntax of the FLUCQ command* Activation:: Starting FlucQ from a CHARMM input file* Charge solution:: Solving for exact charges* Reference energy:: Setting the ``zero'' for FlucQ polarisation* Caveats:: Changes to be aware of; known limitations* Using FlucQ with QM:: Necessary changes for use with CADPAC or GAMESS* Examples:: Simple uses of the FLUCQ command* Implementation:: Mathematical and computational details#File: flucq, Node: Syntax, Next: Activation, Prev: Top, Up: Top [SYNTAX FLUCq] FLUCq { ON init-spec (atom selection) } { OFF } { PRINt } { EXACt exac-spec } { REFErence { GAS exac-spec } } { { SOLVent exac-spec } } { { CURRent } } { { ENERgy real } } DYNAmics ... thermo-spec init-spec::= [GROUp] [NOFIxed] exac-spec::= [TIMEstep real] [ZETA real] [TQDEsired real] [PRINt] thermo-spec::= [FQTEmp real] [FQUNit integer] { FQTCoupling real } ! weak coupling { FQMAss real nose-spec } ! Nose-Hoover { FQSCale integer } ! velocity scaling nose-spec::= [FQTOlerance real] [FQITerations integer]#File: flucq, Node: Activation, Next: Charge solution, Prev: Syntax, Up: Top FlucQ code is enabled within CHARMM by means of the FLUCQ ONcommand. Future energy calculations will then include an extra energyterm - FQPO, the FlucQ polarisation energy, while dynamics simulationsinvolve a new energy property - FQKI, the FlucQ charge kinetic energy.Once FlucQ is active, the selected atoms are treated as extra degreesof freedom, free to fluctuate under the charge forces in the system,and, by assigning each atom type a fictional charge "mass", thesecharges can be accelerated in a conventional dynamics simulation, in acompletely analogous way to the Cartesian degrees of freedom. If atoms are selected by the FLUCQ command which cannot be modelled(i.e. they are QM atoms, or have no FlucQ parameters defined for them)they will be automatically removed from the selection. The FlucQ polarisation energy, FQPO, is an intramolecularinteraction; in full electronegativity equalisation, every atominteracts through space, by means of a modified Coulomb-typeinteraction, with every other atom in the molecule. In thisimplementation, the only interactions calculated are those alongdefined CHARMM bonds (even those with zero force constants). [GROUp] conserves charge within groups, rather than the defaultbehaviour of conserving charge within residues; this prohibits chargetransfer between groups. Note that the FlucQ model makes no restrictionon the degree of charge transfer within each residue or group, or thedistance over which this transfer can occur. [NOFIxed] instructs FlucQ that some or all of the bond lengthsbetween FlucQ-selected atoms are free to change during a simulation.This forces the FlucQ code to recalculate the intramolecularinteraction at each step; since this is a costly calculation, thedefault is to use interactions parameterised for equilibrium bondlengths, with which it is strongly recommended to combine constraintmethods such as SHAKE BONH PARA. The FLUCq PRINt command simply prints the current values of allcharges and charge forces (from the last energy calculation). A similareffect can also be achieved with the standard SCALAR command (seescalar.doc for information on other FlucQ parameters available with theSCALAR command). The FLUCQ OFF command disables the FlucQ code. Further energycalculations will not include FlucQ terms. Note, however, that if thecharges have been modified by FlucQ, they will remain at their alteredvalues. Default behaviour during dynamics is to allow the charge degrees offreedom to fluctuate freely; however they can be thermostatted at a givencharge "temperature" by passing extra options to the DYNAmics command:- [FQTEmp <real>] specifies the charge temperature (default 0). [FQTCoupling <real>] (default 0) if set, uses the Berendsen weakcoupling algorithm to thermostat the charges. The coupling parameteris given in 1/ps, and is analagous to the TCONS/TCOU dynamics options. [FQMAss <real>] (default 0) if set, uses Nose-Hoover thermostatting,with the given mass. The tolerance of the Nose-Hoover iterations can beset with FQTOlerance (default 1.0d-7), and the maximum number ofiterations with FQITerations (default 100). Thermostatting parameters (number of iterations, scale factor, etc.)can be written out to a given unit number at every dynamics step by usingthe FQUNit (default -1: no write) option. [FQSCal <integer>] (default 0) if set, performs simple charge velocityscaling every FQSCal dynamics steps.The initialization process dimensions FlucQ with the current stateof the system. The QM region, if any, is detected, and the FlucQ atomselection will then interact with the QM region. Thus, the FLUCQcommand should be placed after any QUANTUM, CADPAC, or GAMESS command,and if the total number of atoms in the system is modified, FlucQshould be disabled prior to this change and reinitialized afterwards. To skip FlucQ energy calculations entirely, use the SKIP FQPOL FQKINcommand. The QM/MM FlucQ interaction is calculated in line with thestandard QM/MM electrostatic interaction, and as such is suppressedwith the SKIP QMEL command. Finally, the intermolecular