2 revisions of this difference (71086and 79692) were not found.
This is usually caused by following an outdated diff link to a page that has been deleted. Details can be found in the deletion log.
Revision as of 22:51, 7 January 2009
usage: hlamc.pl [options] PDBfile options: [-n runs] [-seq file] [-par mctemp=value,mode=coupled|free, [no]keepchain,[no]eval, caforce=value,hmcmforce=value] [-latpar ncycle=value,icycle=value,temp=value] [-aapar CHARMMparams] [-evalpar CHARMMparams] [-dir name] [-enstag name] [-log file] [-evallog file] [-aalog file]Show source
This utility is used for running single hybrid lattice/all-atom simulations. A replica exchange
version <docmark>hlamcrex.pl</docmark> is also available.
Two Monte Carlo simulation schemes are available for hybrid simulations that combine lattice and all-atom models. In both modes lattice models are effectively simulated on an all-atom surface combining the advantages of reduced lattice models with a more accurate all-atom potential energy function.
In the first mode, called the "free" mode, a Monte Carlo simulation cycle consists of a lattice simulation and all-atom reconstruction of the final lattice chain followed by all-atom minimization and molecular dynamics. The all-atom energy at the end of each cycle is used for accepting or rejecting conformations according to Metropolis criteria.
In the second mode, called "coupled", lattice and all-atom simulations are more tightly coupled. Here, final conformations from very short lattice simulations are used for generating side chain center of mass restraints in continuous all-atom minimization or molecular dynamics runs. As in the first mode, all-atom energies are used for accepting or rejecting conformations in a Monte Carlo scheme. The advantage of the second method is that the all-atom structures remain mostly relaxed throughout the simulation while following the guiding restraints from the lattice simulations. However, one faces the same barriers due to steric hindrance as in conventional all-atom simulations that are easily overcome with the reduced lattice model in the first mode.
The default mode is "free". It can be selected with the mode keyword of the -par option. Other parameters are the Metropolis 'temperature' in units of kcal/mol that is set with mctemp, the force constant that is used for C-alpha restraints in the all-atom part during the "free" mode, the force constant used in the "coupled" mode for side chain center restraints, and two additional options [no]keepchain and [no]eval. The first one sets whether lattice simulations are continued from the final chain of the previous lattice run (keepchain, default) or from a chain generated at the end of the all-atom simulation section (nokeepchain). The second option controls whether a separate instance of CHARMM is used for evaluating all-atom energies at the end of the simulation cycle(eval vs. noeval). This is useful when a different, more costly function, e.g. using an implicit solvation term, should be used for evaluating energies while a simple vacuum energy function may be sufficient for restrained all-atom minimizations and short molecular dynamics runs. The number of total Monte Carlo cycles is set with -n.
Lattice simulations require a MONSSTER sequence file as input with -seq. Other lattice simulation parameters such as the number of lattice simulation Monte Carlo cycles, the internal temperature or potential component scaling factors as listed in <docmark>geninput.pl</docmark> are given with -latpar.
All-atom simulation parameters are set with -aapar. The complete list of possible parameters is given <A HREF="http://mmtsb.scripps.edu/cgi-bin/charmmpar" TARGET=cpar>here</A>. Minimization runs are performed if sdsteps and/or minsteps are set to non-zero values. Molecular dynamics is turned off by default but will be performed if dynsteps is set. Parameters for a separate energy evaluation are set with -evalpar.
All output is written to the current directory unless a different directory is specified with -dir. Structures that are accepted at the end of a Monte Carlo cycle are saved into an ensemble under the same directory. The tag that is used can be set with -enstag.
Log files for the all-atom simulation part, all-atom evaluation, and overall hybrid sampling are generated if the options -aalog, -evallog, or -log are used.
- usage information
hlamc.pl -n 10 -dir data -seq 1vii.seq -par mctemp=30,mode=free,noeval -latpar ncycle=2,temp=1.2 -aapar dielec=rdie,epsilon=4.0,sdsteps=20,minsteps=100 -aalog aa.log -log - 1vii.exp.min.pdb
runs hybrid lattice sampling on an all-atom energy surface. The lattice simulations are run at a temperature of 1.2 internal units for 2 MC cycles. All-atom structures are then rebuild and minimized with a distance-dependent dielectric. No separate copy of CHARMM is run for energy evaluation. The Monte Carlo temperature is set to 30.0. A log file is produced for the all-atom minimization runs and hybrid sampling output is written to standard output.
## hlamc ## mcRun 1 etrial: 134.622850 accepted ## hlamc ## mcRun 2 etrial: 169.625670 rejected ## hlamc ## mcRun 3 etrial: 118.470870 accepted ## hlamc ## mcRun 4 etrial: 198.314910 rejected ## hlamc ## mcRun 5 etrial: 171.513970 rejected ## hlamc ## mcRun 6 etrial: 200.561280 rejected ## hlamc ## mcRun 7 etrial: 163.765020 rejected ## hlamc ## mcRun 8 etrial: 177.650390 rejected ## hlamc ## mcRun 9 etrial: 203.281110 rejected ## hlamc ## mcRun 10 etrial: 193.945440 rejected
hlamc.pl -n 10 -dir data -seq 1vii.seq -par mctemp=2,mode=coupled,nokeepchain -latpar ncycle=2,temp=1.2 -aapar dynsteps=500,dielec=rdie,epsilon=4.0,sdsteps=20,minsteps=100 -evalpar gb,nocut -evallog eval.log -aalog aa.log -log - 1vii.exp.min.pdb
runs coupled-mode hybrid lattice/all-atom sampling for 10 cycles in the directory data. The lattice simulations are run at an internal temperature of 1.2 for 2 MC cycles. The all-atom part consists of 20 steps steepest descent minimization starting from the rebuild lattice chain followed by 100 steps adopted-basis Newton-Raphson (default) minimization and 500 steps of molecular dynamics, all with a distance dependent dielectric and an epsilon value of 4. At the end of each run the energies are evaluated with a Generalized Born term and no cutoffs. The evaluation temperature for the Monte Carlo step is set to 2. Log files are produced for the all-atom simulation and the all-atom energy evaluation. Progress information of the hybrid sampling is written to standard output.
## hlamc ## mcRun 1 etrial: -1114.734410 accepted ## hlamc ## mcRun 2 etrial: -1105.549000 rejected ## hlamc ## mcRun 3 etrial: -1111.145280 rejected ## hlamc ## mcRun 4 etrial: -1114.198550 rejected ## hlamc ## mcRun 5 etrial: -1104.192360 rejected ## hlamc ## mcRun 6 etrial: -1107.517300 rejected ## hlamc ## mcRun 7 etrial: -1093.655610 rejected ## hlamc ## mcRun 8 etrial: -1106.020790 rejected ## hlamc ## mcRun 9 etrial: -1107.190930 rejected ## hlamc ## mcRun 10 etrial: -1105.878440 rejected