CADD picture


This page was originally developed by Kenno Vanommeslaeghe here.

MacKerell Lab
Computer-Aided Drug Design Center

University of Maryland, Baltimore
Baltimore, MD 21201

CGenFF logo


Top - Back to CGenFF main page

Generating atom types, charges and parameters by analogy

  1. Register. You will receive an e-mail with information on how to activate your account. Presently, only academic and government users are allowed.
  2. Create a mol2 file of your molecule, being careful that that all hydrogens are present, in the correct protonation and tautomeric states, and that the bond orders are correct. Unfortunately, there are few open source molecular editors that allow the user to visualize and edit bond orders and save the result as a valid mol2 file that is conform the mol2 standard. Here are 2 closed-source programs that are free of charge and allow you to do so: Additionally, here are some non-free programs (other than the paid versions of the above): Please email Kenno Vanommeslaeghe if you think additions should be made to these lists! In particular, it would be nice to know of an open-source program that has the same abilities.
    Note that mol2 files from the Zinc database are generally high-quality.
  3. Upload the resulting mol2 file. Note that CGenFF should not be used for biological macromolecules; see the notes on the main page or the FAQ for details.
  4. If the interface returns warnings, look at them carefully - warnings often imply that the parameters may not be sufficiently accurate for use without validation and/or optimization.
  5. Download the str file and open it in a text viewer or editor. In particular, look at the "penalty scores" associated with the partial charges and parameters. Penalties between 10 and 50 indicate that some basic validation is recommended; penalties higher than 50 usually are usually associated with parameters or charges that need additional optimization. It should be noted that even in cases were the penalty score is low, the accuracy of the parameters cannot be 100% guaranteed. In all cases it is recommended that the user perform some level of validation of the parameters prior to use.
  6. After validation and/or optimization of the high-penalty parameters and charges, the molecule is ready for MD simulations. Just remember that when writing a paper using CGenFF parameters, the versions of both the CGenFF force field and the CGenFF program should be mentioned in order to ensure reproducibility!
Note: frequent users may wish to obtain a binary license. The procedure for obtaining a free-of-charge not-for-profit license is initiated by e-mailing us; it may take up to a few weeks and will require someone with signature authority at your institution to sign a license agreement that needs to be sent back to us. For-profit users may obtain the CGenFF program from SilcsBio, LLC.

Contents - Back to CGenFF main page

Running CHARMM simulations with the resulting parameters

Prerequisite: a CHARMM version that supports CGenFF; see the FAQ for details.

Basic energy minimization

  1. Download the main CGenFF topology and parameter files.
  2. Verify that your stream file is compatible with the CGenFF version you downloaded (all the files contain comments with version information). Using incompatible files may not result in errors but may give rise to inconsistent energetics! If needed, older CGenFF versions can be found on the download page. (Currently, the available version of the CGenFF program is 1.0.0, which requires the latest version of the CGenFF force field, 3.0.1 .)
  3. Rename your stream file to molecule.str and your mol2 file to molecule.mol2 , where molecule is the name of your molecule as found in the RESI line in your stream file.
  4. Download the minimize.inp CHARMM script and the mol2crd shell script and save them in the same directory where your stream file and your CGenFF topology and parameters files are located.
  5. Run the following commands:
    chmod 755 mol2crd    (this needs to be run only once)
    charmm resi=molecule -i minimize.inp -o molecule.out
  6. If everything went well, molecule.out should have "NORMAL TERMINATION BY END OF FILE" somewhere in the last 20 lines. If not, go to the beginning of the file and search for occurrences of the word "WARNING". In 90% of cases, the first warning message is the cause of the problem.
  7. If CHARMM finished successfully, the minimized structure is saved to molecule_min.pdb and molecule_min.crd . It is recommended to visualize this final geometry; if it is not reasonable, something may be wrong with the force field (although a bad minimized structure may also result from having an unreasonable geometry in the initial mol2 file, as reproduced in molecule_init.pdb and molecule_init.crd ). Note that a good minimized structure is not a guarantee that the force field is good enough for all purposes; in particular, a minimized structure in vacuum is of limited relevance for assessing intermolecular interactions.

MD simulation in the presence of a protein and explicit water

A discussion on how to run MD simulations is beyond the scope of the present document; we only wish to remind our users to (1) download and unpack the latest CHARMM force field distribution and (2) overwrite the main CGenFF topology and parameter files downladed in the previous step with the latest CGenFF version (the CGenFF program works only with the latest CGenFF release and CGenFF releases are often ahead of the releases of the CHARMM force field). Instructions on how to read in the main protein topology and parameter file together with CGenFF can be found on the main page or in the FAQ. After reading both force fields, the stream file generated by the CGenFF program can be read as demonstrated in minimize.inp . More general information on how to run MD simulations with CHARMM can be found in the CHARMM tutorial, CHARMM-GUI, CHARMMing, and the "script archive" section of the CHARMM forums.

Using CHARMM's Domain Decomposition (DOMDEC) or implicit solvents

The DOMDEC feature strongly improves CHARMM's parallelization efficiency. However, it requires the molecule to be divided into spatially localized groups. It shares this prerequisite with most of CHARMM's implicit solvent models. Therefore, we offer a program to insert GROUP statements into any toppar stream file generated by the CGenFF program. Note that groups will not necessarily have integer charges, which precludes their use with CHARMM features that preform multipole expansion on them. Fortunately, this is not the case for DOMDEC nor for any of the implicit solvent models we know of. The script is invoked as follows:
regroup.awk molecule.str > molecule_regrouped.str

Contents - Back to CGenFF main page

Running simulations with other programs


The following steps should allow for CGenFF to be used with NAMD:
  1. Download and unpack the latest CHARMM force field distribution
  2. Overwrite the main CGenFF topology and parameter files downladed in the previous step with the latest CGenFF version (the CGenFF program works only with the latest CGenFF release and CGenFF releases are often ahead of the releases of the CHARMM force field)
  3. Download toppar_water_ions_namd.str and put it alongside the other parameter files
  4. Generate the system of interest in CHARMM and write a PSF in XPLOR format
    - OR -
    Generate a PSF using VMD's psfgen plugin (note that we haven't tested this)
    - OR -
    Generate your system and its XPLOR PSF with CHARMM-GUI (yes, it does officially support NAMD)
  5. ONLY FOR VERY OLD NAMD VERSIONS: edit the main CGenFF parameter file (ie. par_all36_cgenff.prm ) and remove the ATOMS section with MASS entries.
  6. Read the XPLOR PSF and all the parameter files (plus the str file with molecule-specific parameters) into NAMD as follows:
    structure system-xplor.psf
    paraTypeCharmm on
    parameters par_all36_prot.prm
    mergeCrossterms yes
    parameters par_all36_lipid.prm
    parameters par_all36_carb.prm
    parameters par_all36_cgenff.prm
    parameters molecule.str
    parameters toppar_water_ions_namd.str
Note: for protein-ligand complexes, combined PSF files that include all the molecules, including solvent, are required.


See the multiscale lab's CHARMM parameter assignment protocol worksheet.


There is a GROMACS conversion program for ParamChem-generated CGenFF toppar stream files. The comments section at the beginning of the Python script provides usage information. We will post further instructions soon.
Under Construction icon

Other programs

In principle, CGenFF can be used with any program that supports the CHARMM potential energy function and combining rules, and has switching functions for the nonbonded cutoffs that are suitable for use with the CHARMM force field. In short, if a simulation program has good support for the CHARMM force field, it should not be prohibitively difficult to use CGenFF with it. If you are part of the community around one of these other programs, and you have a protocol online for using CGenFF with your program that is well-validated and follows best practices, please contact Kenno Vanommeslaeghe or you can use the ParamChem contact address.

Contents - Back to CGenFF main page

How it works

The atom typer is rule-based and programmable, making it easy to implement complex atom typing rules and to update the atom typing scheme as the force field grows. Assignment of bonded parameters is based on substituting atom types in the definition of the desired parameter. A penalty is associated with every substitution and the existing parameter with the lowest total penalty is chosen as an approximation for the desired parameter. Charges are assigned using an extended bond-charge increment scheme that is able to capture inductive and mesomeric effects across up to 3 bonds. More details can be found in the following articles:
K. Vanommeslaeghe, A. D. MacKerell Jr., J. Chem. Inf. Model. 2012, 52, 3144-3154.
K. Vanommeslaeghe, E. P. Raman, A. D. MacKerell Jr., J. Chem. Inf. Model. 2012, 52, 3155-3168.

Contents - Back to CGenFF main page


This material is based upon work supported by the National Science Foundation under Grant No. 0823198. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Contents - Back to CGenFF main page

Last updated Friday, the 17th of April 2015