|
|
| From: |
"David Reichert" <reichertd -8 at 8- mirlink.wustl.edu> |
| Date: |
23 Aug 1997 15:41:20 -0500 |
| Subject: |
Summary Parameterization Tools |
************************************************
* David E. Reichert, Ph.D. *
* Mallinckrodt Inst. of Radiology *
* Div. Radiological Sciences, *
* Washington University School of Medicine *
* 510 S. Kingshighway Blvd., Campus Box 8225 *
* St. Louis, MO 63110 *
* *
* phone: (314)362-8461 *
* fax: (314)362-9940 *
* e-mail: reichertd -8 at 8- mirlink.wustl.edu *
************************************************
<<<<<< Attached TEXT file named "CCL Parameterization Tools" follows >>>>>>
Almost a month ago I posted a query:
>From: David Reichert
>Date: Tue, Jul 29, 1997 1:19 PM
>Subject: CCL:Parameterization Tools
>To: CCL post
>
>I'm looking for some info and/or code to help automate the development of
>forcefield parameters, hopefully the list readers can provide some leads.
>Given a set of x-ray or NMR structures and/or results from ab initio
>calculations are there programs that can automate the development of
>parameters for a given forcefield, ie eq bondlengths, angles, stretching
>constants etc... Any pointers would be greatly appreciated. (I'm getting
tired
>of doing this by hand, there must be a better tool somewhere !)
>thanks,
>Dave
The following is a summary of replies that I received. I'd like to thank
everyone who took the time to send me some leads.
--------------------------------------------------------------------------------------
From: dalke ^at^ mag.com
Date: Tue, Jul 29, 1997 1:39 PM
Subject: Re: CCL:Parameterization Tools
To: David Reichert
> Given a set of x-ray or NMR structures and/or results from ab initio
> calculations are there programs that can automate the development of
> parameters for a given forcefield, ie eq bondlengths, angles, stretching
> constants etc
X-PLOR can do this. See the man page at:
http://xplor.csb.yale.edu/manual/htmlman/node56.html
Andrew Dalke
dalke()at()mag.com
From: Richard Wisniewski
Date: Tue, Jul 29, 1997 1:43 PM
Subject: Re: CCL:Parameterization Tools
To: David Reichert
At your university the person to contact on this subject might be Dr Jay
Ponder...
Richard Wisniewski
NASA ARC
From: Dr. Mike Gilson
Date: Tue, Jul 29, 1997 1:59 PM
Subject: force field
To: David Reichert
Hi,
I am not sure whether there is code to to this, but the following
papers might conceivably be of interest....
good luck,
Mike
====================================
author = "J. R. Maple and {M.-J.} Hwang and T. P. Stockfisch and
U. Dinur and M. Waldman and C. S. Ewig and A. T. Hagler",
title = "Derivation of Class {II} Force Fields. 1. {M}ethodology
and Quantum Force Field for the Alkyl Functional
Group and Alkane Molecules",
OPTcrossref = "",
OPTkey = "",
journal = "J. Comput. Chem.",
year = "1994",
volume = "15",
OPTnumber = "",
pages = "162-182",
OPTmonth = "",
OPTnote = "",
OPTannote = ""
}
()at()Article{cff2,
author = "{M.-J.} Hwang and T. P. Stockfisch and A. T. Hagler",
title = "Derivation of Class {II} Force Fields. 2.
{D}erivation and Characterization of a Class {II}
Force Field, {CFF}93, for the Alkyl Functional Group
and Alkane Molecules.",
OPTcrossref = "",
OPTkey = "",
journal = "J. Am. Chem. Soc.",
year = "1994",
volume = "116",
OPTnumber = "",
pages = "2515-2525",
OPTmonth = "",
OPTnote = "",
OPTannote = ""
}
()at()Article{cff3,
author = "J. R. Maple and {M.-J.} Hwang and T. P. Stockfisch and
A. T. Hagler",
title = "Derivation of Class {II} Force Fields. 3.
{C}haracterization of a Quantum Force Field for Alkanes.",
OPTcrossref = "",
OPTkey = "",
journal = "Isr. J. Chem.",
year = "1994",
volume = "34",
OPTnumber = "",
pages = "195-231",
OPTmonth = "",
OPTnote = "",
OPTannote = ""
}
From: Andrew Rohl
Date: Tue, Jul 29, 1997 6:45 PM
Subject: Re: CCL:Parameterization Tools
To: David Reichert
Dr Julian Gale has a program GULP which can do this. His email is
julian' at \`ri.ac.uk
Andrew
From: Per-Ola Norrby
Date: Wed, Jul 30, 1997 2:27 AM
Subject: Re: CCL:Parameterization Tools
To: David Reichert
Dear David,
I feel I must respond to your question. I do have such a tool, and
I plan to give it away for free, but I must wait to have it accepted. I
submitted it in May, so I hope to have some response soon.
The reason I answer before I have something I can share is that
what you requested is EXACTLY what I've been developing. There are some
similar tools available already, but most are limited to one kind of data.
I have made sure that I can include most kinds of experimental and
theoretical results in the parameterization.
I have been working with MacroModel, for those force fields my
routines work without modification, but I've started to modify the routines
for other packages that run under Unix. The main requirement is that you
should be able to run the program from a Unix script, and generate a text
output containing the force field predicted data. In short, what I do is
predict bond lengths, angles, dihedrals, energies, cartesian energy
derivatives etc. from the force field, compare it to the reference data,
and automatically vary the parameters to obtain the best possible
correspondance in a least squares sense. It's a bit tricky to find scale
factors to allow the routines to treat all different kinds of data on an
equal basis, but I have a working solution (which I continually modify).
As for already existing tools, there are a few (I don't have
experience with them, I've just read about them):
* The BIOSYM PROBE program will allow you to optimize parameters to fit
quantum chemical energies. It seems VERY efficient, and has been used
successfully by both Hagler et al. (CFF), and Halgren et al. (MMFF).
However, I believe you must be a member of the consortium to get access to
it. See: A. T. Hagler et al., J. Comput. Chem., 15, 162 (1994). A. T.
Hagler et al., J. Am. Chem. Soc., 116, 2515 (1994). T. A. Halgren, J. Am.
Chem. Soc., 114, 7827 (1992). T. A. Halgren, J. Comput. Chem., 17, 490
(1996) and following papers in the same issue.
* There are at least two programs I know about where you can define your
own force field (including the functional form) and, I believe,
automatically optimize the parameters. These are PEFF from Jan Dillen (J.
L. M. Dillen, J. Comput. Chem., 13, 257 (1992)), and TINKER from Jay Ponder
(http://dasher.wustl.edu/).
There are several other published methods, but I'm not aware of any
program that goes with them. Some are based on the original CFF by Lifson
and Warshel, and there may be a program available from the Lyngby group (K.
Rasmussen et al., Acta Chem. Scand., 48, 548 (1994) and following paper).
I've started using my routines with TINKER and with MM3(94), it
seems to work, but there are still a few bugs.
Best Regards,
Per-Ola Norrby
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
* Per-Ola Norrby, Research Associate
* The Royal Danish School of Pharmacy, Dept. of Med. Chem.
* Universitetsparken 2, DK 2100 Copenhagen, Denmark
* tel. +45-35376777-506, +45-35370850 fax +45-35372209
* Internet: peon ( ( at ) ) medchem.dfh.dk, http://compchem.dfh.dk/
Message for David Reichert
From: szilagyi ( ( at ) ) indy.mars.vein.hu
Date: Wed, Jul 30, 1997 3:02 AM
Subject: Re: CCL:Parameterization Tools
To: David Reichert
Dear David,
we have developed a fully automatized parameterisation algorithm, which
is going to be published in a couple of weeks. The method performs a
statistical analyses of available structural sources and then an
electronic sheet guided stepwise parameter adjustement is taken place.
If you are interested in more details please let me know about your
current research problem!
Best wishes,
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Robert K. Szilagyi Mueller Laboratory
ph.d. student Dept. Organic Chemistry
University of Veszprem
szilagyi at.at mm2.vein.hu Veszprem, H-8201
POB. 158; Egyetem u. 10.
Phone: +36 88 422022 extn. 395 HUNGARY
Mobile: +36 20 461413
Fax: +36 88 427492 http://mm1.vein.hu/
======================================================================
********** All opinions are my own, NOT my employer's ! **********
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: szilagyi' at \`indy.mars.vein.hu
Date: Mon, Aug 4, 1997 1:03 AM
Subject: Re: CCL:MM parameter development
To: David Reichert
Dear David,
an automatic parameter optimisation tool was developed in our
laboratory. We have tested on alkylphosphines (published in
J.Mol.Struct. but not yet rolled out), alkylsilanes, ruthenium and
tungsten organometallic complexes. It works fine and converges rapidly.
The method will be published in J.Comput.Chem. very soon!
The method developed can take into account any structural consideration
based on X-ray, IR, Electron diffraction of microwave spectroscopy.
If you are interested in more details let me know!
Best wishes,
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Robert K. Szilagyi Mueller Laboratory
ph.d. student Dept. Organic Chemistry
University of Veszprem
szilagyi -x- at -x- mm2.vein.hu Veszprem, H-8201
POB. 158; Egyetem u. 10.
Phone: +36 88 422022 extn. 395 HUNGARY
Mobile: +36 20 461413
Fax: +36 88 427492 http://mm1.vein.hu/
======================================================================
********** All opinions are my own, NOT my employer's ! **********
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: Brian Teppen
Date: Wed, Jul 30, 1997 12:50 PM
Subject: CCL:Parameterization Tools -Reply
To: David Reichert
Hi, Dave:
I have used a tool provided by Kjeld Rasmussen of the Technical University of
Denmark, called CFF. He studied with Lifson way back around 1970 and has been
developing Lifson's program as a parameter-fitting tool, rather than as a
macromolecular simulator, as others have done.
The program is a mix of Fortran for the actual nonlinear least-squares
optimization and C for a graphical interface that allows one to monitor the
progress of the parameter optimization. One can input molecular structures and
molecular crystals, and one can optimize to structures, unit cell parameters,
and vibrational frequencies simultaneously.
So far, Dr. Rasmussen has given the source code to those who ask. He has an
ftp server with a tar file that includes a user manual. You can contact him at
kjr %-% at %-% kemi.dtu.dk
Program descriptions and optimization examples can be found in:
Engelsen, S. B.; Fabricius, J.; Rasmussen, K. Acta Chem. Scand. 1994, 48,
548-552; 553-565.
Jonsdottir, S. O.; Rasmussen, K. New. J. Chem. 1995, 19, 1113-1122.
Fabricius, J.; Engelsen, S. B.; Rasmussen, K. New. J. Chem. 1995, 19,
1123-1137.
Niketic, S. R.; Rasmussen, K. The consistent force field: A documentation;
Lecture Notes in Chemistry 3; Springer-Verlag: Heidelberg, FRG, 1977.
Rasmussen, K.; Engelsen, S. B.; Fabricius, J.; Rasmussen, B. In Recent
experimental and computational advances in molecular spectroscopy. NATO ASI
Series C: Mathematical and Physical Sciences; Fausto, R., Ed.; Kluwer Academic
Publishers: Dordrecht, 1993; Vol. 406, pp. 381-419.
Best wishes,
Brian J. Teppen teppen -AatT- srel.edu
Advanced Analytical Center for Environmental Sciences
Savannah River Ecology Laboratory
University of Georgia
Drawer E
Aiken, SC 29802
phone:803-725-8157 fax:803-725-3309
Message for David Reichert
From: Dayong He
Date: Fri, Aug 1, 1997 1:46 PM
Subject: Re: CCL:MM parameter development
To: David Reichert
Hi, David,
Have you tried the scientists in MSI, there are several guys very
expertise in developing force field.
Dayong
_____________________________________________________________________________
Dayong He
Department of Chemistry
Rutgers, The State University of New Jersey
Piscataway, NJ 08855
Tel: (732)445-4619(o), Email: yong : at : rutchem.rutgers.edu
_____________________________________________________________________________
Message for David Reichert
From: Donald E. Williams
Date: Tue, Aug 19, 1997 11:58 AM
Intermolecular force field development with npg
Program nbp (for nonbonded potentials) is a software tool
which derives an optimized intermolecular force field from molecular
crystal and/or molecular cluster data. The input to the program is
a set of structures encoded in the mpa/mpg (for molecular packing
analysis/molecular packing graphics) format. All atoms (or sites)
in different molecules are assumed to interact via (exp-6-1) or
(n-6-1) nonbonded potentials. Atoms of a given element can be
classified into several potential types. The effect of net atomic
(or site) charges is included in the force field. The force field
can be scaled to one or more energies such as heats of sublimation or
heats of association.
The program can be used to find a customized force field for
any training subset of structural and energy data, find a general
force field for a large training set of data, or to extend an
existing force field to include different types of atoms. The output
from the program is an optimized intermolecular force field which best
represents the training set data.
Mpa/mpg is required for proper operation of nbp. A description
of mpa/mpg is appended. Net atomic charges may be obtained with pdm97,
which is also described.
Brief history of our approach to intermolecular force field development
Early efforts to produce intermolecular force fields began
with graphite and noble gases. Crowell (1958) derived a C...C
potential from graphite using the interlayer energy, interlayer
distance, and compressibility. An analogous procedure was used by
Williams (1972a) to find (exp-6) potentials for Ne, Ar, Kr, and Xe from
crystal structure, heat of sublimation, and compressibility data.
Kitaigorodsky's book (1961), "Organic Chemical Crystallography",
stimulated a great deal of interest in the packing of organic molecules,
especially hydrocarbons. He later published (1973) another treatise
on the same subject.
An early effort produced a quantitative nonbonded force field
for aromatic hydrocarbon crystals (Williams, 1966). This paper was
much cited and was designated a citation classic by the Institute for
Scientific Information, publishers of the "Citation Index". The
hydrocarbon work was extended to include nonaromatics (Williams, 1967)
and net atomic charges (Williams, 1974). The overall method was
described as "molecular packing analysis" (Williams, 1972b).
The method for derivation of intermolecular force fields was
further refined in connection with a study of perchlorohydrocarbon
crystal structures (Hsu & Williams, 1980). Nonbonded potential
functions were derived for nitrogen in azahydrocarbons (Williams & Cox,
1984), oxygen in oxohydrocarbons (Cox, Hsu, & Williams, 1981),
fluorine in perfluorohydrocarbons (Williams & Houpt, 1986), chlorine
in perchlorohydrocarbons (Hsu & Williams, 1980), and chlorine in
dichlorine (Williams & Gao, 1997a). The current version of nbp was
used to derive a force field for boron and hydrogen in crystalline
boranes (Williams & Gao, 1997b).
The number of workers and publications in this field has greatly
expanded. There are now available a variety of intermolecular force
fields from various research groups. Often the intermolecular force
field was not the primary goal, but it was developed in connection
with an intramolecular force field. Discussion and references to
recent work can be found, for example, as articles in the series
"Reviews in Computational Chemistry", in the Journal of Computational
Chemistry, and other major journals. Proprietary force fields also
have emerged where sometimes the basis for establishment of the
parameters is not clearly stated.
References
Crowell, A. D.(1958). Potential Energy Functions for Graphite. J. Chem.
Phys. 29, 446.
Cox, S. R.; Hsu, L. Y.; Williams, D. E. (1981). Nonbonded Potential
Function Models for Crystalline Oxohydrocarbons. Acta Cryst. A37, 293.
Hsu, L. Y.; Williams, D. E. (1980). Intermolecular Potential Function
Models for Crystalline Perchlorohydrocarbons. Acta Cryst. A36, 277.
Kitaigorodsky, A. I. (1961). Organic Chemical Crystallography.
Consultants Bureau, New York.
Kitaigorodsky, A. I. (1973). Molecular Crystals and Molecules. Academic
Press, New York.
Williams, D. E. (1966). Nonbonded Potential Functions Derived from
Crystalline Aromatic Hydrocarbons. J. Chem. Phys. 45, 3770.
Williams, D. E. (1967). Nonbonded Potential Functions Derived from
Crystalline Hydrocarbons. J. Chem. Phys. 47, 4680.
Williams, D. E. (1972a). Direct Calculation of Thermal Expansion and
Molecular Reorientation from Nonbonded Potential Anharmonicity and
Thermal Amplitudes. Acta Cryst. A28, 84-88.
Williams, D. E. (1972b). Molecular Packing Analysis. Acta Cryst. A28,
639-635
Williams, D. E. (1974). Coulombic Interactions in Crystalline
Hydrocarbons. Acta Cryst. A30, 71.
Williams, D. E.; Cox, S. R. (1984). Nonbonded Potentials for Aza-
hydrocarbons. The Importance of the Coulombic Interaction. Acta
Cryst. B40, 404.
Williams, D. E.; Gao, D. (1997a). Effects of the Molecular Electric
Potential and Anisotropic Repulsion in the Chlorine Dimer and
Crystalline Chlorine. Inorg. Chem. 36, 782-786.
Williams, D. E. & Gao, D. (1997b). Intermolecular Force Field Parameters for
Boranes. Acta Cryst., Sec. B, submitted.
Williams, D. E.; Houpt, D. J. (1986). Fluorine Nonbonded Potential
Parameters Derived from Crystalline Perfluorocarbons. Acta Cryst.
B42, 286.
Williams, D. E.; Starr, T. L. (1977). Calculation of the Crystal
Structures of Hydrocarbons by Molecular Packing Analysis. Comp. Chem.
1, 173.
For further information contact:
Dr. Donald E. Williams
Department of Chemistry
University of Louisville
Louisville, KY 40292 USA
email: dew01 "-at-" xray5.chem.louisville.edu
Molecular Packing Analysis/Molecular Graphics
Mpa/mpg is a suite of programs designed for molecular packing
analysis and molecular packing graphics in the unix environment. The
capability of the mpa module extends from minimization of the energy of
association of two or more molecules (e.g., molecular clusters and
host-substrate docking) to energy minimization of a molecular assembly
in a crystal lattice of any space group symmetry (e.g., crystal polymorph
prediction). The software is capable of predicting space groups and
reduced cells in which a given molecule may crystallize. Provision for
limited molecular flexibility is made through selected rotations about
intramolecular bonds. Molecular and/or crystal symmetry restrictions
may be selectively applied as desired.
A library of commonly available intermolecular force fields is
included (e.g., WH86, MM85, Biosym, Charmm, Dreiding, etc.), and provision
is made for user supplied force fields. The program selects energy
minimization by OREM (off-ridge eigenvector minimization), SD (steepest
descents), or NR (Newton-Raphson) as appropriate for the model. All first
and second derivatives of the energy are calculated analytically; the
second derivative matrix (hessian) is diagonalized and its eigenvectors
determined. Simulated annealing is included to aid in locating global,
rather than subsidiary, energy minima.
The molecular packing graphics (mpg) module is closely integrated
with the molecular packing analysis module to give users fast, convenient,
and useful graphical visualization of the results.
To obtain an abstract of the program and a listing of example
calculations, send an email request to dew01 -x- at -x-
xray5.chem.louisville.edu.
In summary, here are some principal capabilities of mpa/mpg:
> minimization of nonbonded and hydrogen bonded energy
in molecular clusters or crystals (includes molecular docking)
> ab initio prediction of space group symmetry
> uses (n-6-1) or (exp-6-1) interatomic potentials
> built-in force field library (WH86, MM85, Biosym, CHARMM, etc.)
> Newton-Raphson (NR), steepest descents (SD), and
off-ridge eigenvector minimization (OREM)
> OREM with simulated annealing (OREMWA)
> provision for limited molecular flexibility
> multiple molecules in asymmetric unit
> accelerated convergence of lattice sums
> convenient graphical visualization
Also available:
Pdm97, Least-Squares Fitting of the Molecular Electrostatic Potential
With Net Atomic Charges and/or Multipoles
Reliable net atomic charges/multipoles can be found by fitting the
molecular electric potential with program pdm97. The program provides a
choice of geodesic(1), Connolly(2), cubic(3), or user specified grid points
for the electric potential. The program includes calculation of hyperbolic-
restricted charge magnitudes(4). In addition to net atomic charges, the
program also allows any combination of atomic dipoles/quadrupoles, bond
dipoles, as well as the addition of lone pair electron sites if required.
The program directly reads output files from Gaussian-92 and Gaussian-94
programs, and will accept edited input from any program which produces a
molecular electric potential on a grid. Program pdm97 is the most complete
and versatile software package available for this type of calculation.
Molecular interactions occur during host-substrate docking, cluster,
and crystal formation - whenever molecules associate with one another. The
energy and geometry of molecular association is determined by the force
field. As a component of the force field, an accurate set of net atomic
charges is required. Or, if higher accuracy is needed, pdm97 can make
extensions to the net atomic charge model in a variety of ways.
A particularly useful feature of the program is the easy and
transparent way in which fixed charges and charge dependency conditions are
specified. By specifying appropriate charge dependencies (e.g., equal
charges or equal sums of charges), chemical intuition can assist to produce
charges which are transferable between related types of molecules or molecular
ions. Hyperbolic-restricted charges are specifically designed to increase
charge transferability. All charge calculation include a complete error
treatment with standard deviations and correlations between variables.
References
(1) Spackman, M. A., "Potential Derived Charges Using a Geodesic Point
Selection Scheme", J. Comput. Chem 1996, 17, 1-18
(2) Connolly, M. L., "Solvent-accessible Surfaces of Proteins and Nucleic
Acids",
Science 221, 709-713 (1983)
(3) Williams, D. E., "Net Atomic Charge and Multipole Models for the Ab Initio
Molecular Electric Potential", Rev. Comp. Chem. 1991, 2, 219-271.
(4) Bayly, C. L.; Cieplak, P.; Cornell, W. D.; Kollman, P. A., "A Well-Behaved
Electrostatic Potential Based Method Using Charge Restraints for Deriving
Atomic Charges: The RESP Model", J. Phys. Chem. 1993, 97, 10269-10280.
For further information contact Dr. Donald E. Williams, Department of
Chemistry,
University of Louisville, Louisville, Kentucky 40292, USA.
E-mail:dew01 ( ( at ) ) xray5.chem.louisville.edu
Tel:(502)852-5975 Fax:(502)852-8149
Similar Messages
02/21/1996: Summary: CRYSTAL & all
10/13/1998: Parameterization - summary
11/24/1995: New molecular packing software
10/25/1996: charges and potentials
10/02/1996: SUMARY:PD atomic charges
01/20/1997: Software available for LSQ Fit of ESP
04/08/1994: normal coordinate calculation
10/01/1993: torsion of conjugated systems -- summary
01/21/1997: Summary: Electrostatic effects in molecular crystals
11/16/1993: MM Parameters/Macrocycles
Raw Message Text
|
