Next: Keywords Up: Introduction Previous: Installation Contents Index
Subsections
Getting Started
New users to DivCon, or quantum mechanics (QM) in general, are often surprised to learn that artificial atom types and other molecular mechanics constructs are not required to utilize this level of theory. Instead, successful application of QM relies on ``standard'' elements, fairly accurate 3D structures, and CPU time. Further, instead of considering the use of QM as a whole other computational chemistry workflow, the user should simply think of QM as adding another dimension to his or her established, all-atom methods. For instance, in terms of preparation of a Protein Databank (PDB) file for treatment in DivCon, the computational chemist would still need to perform steps such as protonation, basic molecular mechanics ``clean-up'' of any especially troublesome artifacts, and overall formal charge determination. On the other hand, depending upon the quality of the starting structure (and by extension, how much molecular mechanics-based optimization is required), the atomic charges are not important to the QM calculation and the DivCon software can in fact further optimize the structure to a semi-empirical level of theory. Figure 1.1 shows the steps the authors often use to prepare a completely ``raw'' PDB structure and the reader probably has his or her own set of analogous steps.
DivCon Input Files
At its most fundamental, DivCon only requires a list of atoms (elements), the XYZ coordinates of those atoms, residue delimiters for the linear scaling, divide and conquer method, and an overall formal charge. As detailed in the remainder of this manual, the user can then use various keywords to set any non-default parameters - such as the Hamiltonian, an alternate formal charge, optimization parameters, and so on. On the other hand, if the user is happy with the defaults and only desires a single point calculation, he or she is free to simply supply a fully protonated mol2 file (or a list of mol2 files) and let DivCon handle the defaults.
% divcon -i molecule.mol2If the user wants to take maximum advantage of everything DivCon has to offer, a number of additional options are available. These include a native DivCon file format which can take the place of the mol2 file noted above, an optional parameter file to encapsulate the various keywords or parameters of the simulation, and a number of command line switches that can either take the place of a parameter file or supersede the keywords found within that file. Finally, one can also use the included OpenBabel 2.0 plug-in to translate from the other file formats supported by OpenBabel to a ``native'' DivCon input file.
Command Line Arguments
divcon [-i mol.mol2][-p params][-c charge][-opt [?]][-screen][-am1][-f list][-wrtmol2 [?]]
- [-i ?] The -i switch is used to enter a molecule input file. Any of the currently supported formats will do, and the user can provide any number of input files.
- [-t ?] Similar to the -i switch, this option will allow the user to enter a molecule input file. There are two major differences however. First, the molecule that follows the -t will always precede all structures entered using -i and -f. The ``t'' in the switch stands for ``target'' and it is assumed that the structure is an enzyme (note: there are no further error traps!). The second difference between -t and -i is that while -i can accept any number of files, -t can accept only one. One should envision using this switch in conjunction with the -qmscore script noted below.
- [-o ?] During the execution of a multi-job DivCon simulation, the overall success or failure of each job will be recorded in this file. By default, this file will be named divcon_stat.out.
- [-p ?] Read the keywords for the simulation from a parameter file. The keywords within the parameter file take precedence over any found within the native DivCon input file.
- [-c ?] The integral formal charge
of the molecule entered using the -i switch. The charge entered on
the command line takes precedence over the charge within any files.
If more then one file is entered using the -i, -f, or -t switch(es),
the behavior of the -c switch is undefined and the user should instead
use the CHARGE= keyword discussed on page
.
Therefore, the -c is only meaningful when a single structure is being
treated.
- [-s ?] Activate the SHIFT=
keyword (see page
) with the value following
the -s being the starting point for the shift algorithm.
- [-opt [?]] Execute a DivCon
optimization. The -opt switch has four different possible values to
activate different algorithms: lbfgs, conjgrad, bfgs, or steep (see
page
). If the -opt switch is
alone, then the default optimization method is used instead.
- [-screen] Output status data
to screen as the simulation progresses. Equivalent to the SCREEN
keyword detailed on page
.
- [-f ?] Include a list of molecule files. This list can be a mixture any supported file formats and each file in the list should be on its own line. Each file is executed in the order in which entered. If MPI is being used, then the files are distributed to the processors in the set.
- [-wrtmol2 [?]] Output a mol2-formated file including the final XYZ coordinates of the simulation along with the CM1, CM2, or MULLIKEN atomic charges calculated using DivCon. The user can choose which charge model to output by substituting the model for the ? in the switch. The CM2 charges are written by default.
- [-wrtpdb] Output a PDB-formated file enclosing the final XYZ coordinates of the simulation.
- [-wrtin] Output a fully-compliant native DivCon input file with the finally XYZ coordinates and keywords of the simulation.
- [-am1] DivCon supports five different semi-empirical Hamiltonians. Any one of these methods can be chosen on the command line using one of the -am1, -pm3, -mndo, -mndod, or -pddg-pm3 switches. See page
- [-qmscore] Run QMScore algorithm on the files entered using the -t, the -i and the -f switches. The first file on the concatenated list of files (always the file after -t if this switch is provided) is treated as the enzyme while each of the other files in the list is treated as an inhibitor of that enzyme. Note: these inhibitors should already be docked and prepared before treatment using DivCon/QMScore.
- [-pwd] Execute the Pairwise Decomposition method on the entered file(s). This switch would be equivalent to using the PWD keyword.
- [-scrf] Perform all calculations using the self-consistent reaction field (SCRF) algorithm. Using this switch would be equivalent the user adding the SCRF keyword. Note: the DIVPB algorithm will also be employed.
- [-test] DivCon will perform all
of the simulation preparation, license checking, and stop just prior
to actually running any quantum mechanics calculation. This switch
can be used to make sure that the job is ready to go and that it will
not die due to preparation-oriented reasons (such as an open shell
situation). Using this switch is equivalent to including the TESTRUN
keyword discussed on page
.
- [-freq] Perform a frequency calculation
(see page
) on the current conformation
using the FREQ keyword. This command line switch
can be joined with the -opt switch to perform this calculation on
the minimized structure.
The Native DivCon Input File
DivCon supports two major file formats: the ``native'' divcon.in format and the standard mol2 format originally championed by Tripos' Sybyl package. The native format is depicted below where the coordinates, keywords, title, and parameters can be packaged together as a single entity.
At the top of the file are the keywords and the title of the job.
As discussed in section
, most of these
keywords can also be packaged within a separate parameter file if
the user would prefer. Often, the user will want to include simply
the CHARGE= keyword in a single keyword line
and include any other keywords either as command line arguments or
as content within the parameter file. Therefore, the divcon.in file
can be extremely sparse (such as when OpenBabel
is used to generate the file).
The format of atom, coordinate and residue information is:
CHARGE=0 RESIDUE CLUSTER
title
1 O 4.7092732 3.8021271 8.9033802 RES {WAT;O;1;1;1}
2 H 5.0723235 4.3719820 9.5513718 {WAT;H;1;1;1}
3 H 4.9639974 4.1517696 8.0556286 {WAT;H;1;1;1}
4 O 0.8528918 0.3470596 0.1015257 RES {WAT;O;1;2;1}
5 H 0.1648828 0.9302319 -0.1495947 {WAT;H;1;2;1}
6 H 1.6391883 0.8006957 0.2127379 {WAT;H;1;2;1}
END_COORD
CLUSTER
NCORE=1 DBUFF1=4.5 DBUFF2=2.0
END_CLUSTERThe example indicates that the first residue is atom 1-3 and the second residue atom 4-6. Coordinates are in Å and you can use as much white space as you want (maximum characters per line is 80). The {WAT;H;1;1;1} statements are optional as they denote the additional information generally found in a file such as a PDB formatted file, but not required for a QM calculation. The ``WAT'' is the name of the residue while the ``H'' is the PDB atom name. The ``1'', ``1'', ``1'' respectively refer to the residue number, molecule number, and collection number of the species in question. The molecule and collection numbers then correspond to the information provided in the optional CHEMIX/END_CHEMIX parameter block shown below. For the water example above, the ``RES'' delimiters denote where a residue begins and these delimiters are required when the RESIDUE keyword is used.
CHEMIX
COLLECTION 1 2-WATER-MOLECULES
MOLECULE 1 1 1-WATER-MOLECULE-A
MOLECULE 2 1 1-WATER-MOLECULE-B
END_CHEMIXNote: Do not use tabs in the input file. Use space instead.
Parameter File
Instead of packaging the keywords and the parameter blocks within the native DivCon file, the user can prepare ``standard'' simulation sets and execute these sets on any number of coordinate files using a separate parameter file. To build a parameter file, the keywords of the simulation should be placed within a KEYWORD block:
KEYWORD
CLUSTER RESIDUE
END_KEYWORD
CLUSTER
NCORE=1 DBUFF1=4.5 DBUFF2=2.0
END_CLUSTERFollowing the KEYWORD block, the user can place any additional parameter blocks that are needed - such as a CLUSTER block noted above. The native DivCon input file would then only include the total charge (using the CHARGE= keyword), the coordinate information, and the CHEMIX/END_CHEMIX parameter block. Everything else could be moved to the separate parameter file. Once the parameter file is created, it can be loaded on the command line following the ``-p'' switch as discussed in section 1.6.1.1.
Next: Keywords Up: Introduction Previous: Installation Contents Index support@quantumbioinc.com