CHARMM Development Project Forums User Discussion & Questions Parameter Set Discussion Charmm General Force Field (CGenFF) download link

Hi Everyone,

The Charmm General Force Field (CGenFF) for drug-like molecules finally can be downloaded from the MacKerell lab website.

Notes:

• The early alpha version distributed with c36a had some issues and is now obsolete. Only use the version from the website!
• There is a tutorial available for download as well
• I will post any updates to this "discussion topic"

Another note: please never use CGenFF for biological macromolecules!!!
The Protein, Nucleic Acid, Carbohydrate and Lipid force fields are highly optimized for their specific purpose. The general force field is designed to cover any combination of chemical groups. This inevitably comes with a decrease in accuracy for representing any particular subclass of molecules, such as proteins, nucleic acids, carbohydrates and lipids. It is for this reason that we removed the amino acids, nucleotides,... from CGenFF. If you re-introduce them, you will waste a lot of time and obtain bad results.

Here is what you should do instead:

1. Make an "ATOM" section in the protein/nucleic acid/carbohydrate/lipid parameter file, containing all the MASS entries from the corresponding topology file. A sample protein parameter file can be found in the tutorial on the MacKerell lab website.
2. Renumber all "ATOM" fields in the CGenFF topology and parameter files so that they don't coincide with the protein/nucleic acid/carbohydrate/lipid atom numbers. The tutorial contains a script do_remas that does this for you.
3. read in the resulting topology and parameter files as follows:

   
   open unit 10 read form name @TOPPAR/top_all22_prot.rtf
   read rtf card unit 10
   close unit 10
   open unit 10 read form name @TOPPAR/par_all22_prot.prm
   read param card flex unit 10
   close unit 10
   open unit 10 read card name @TOPPAR/top_all36_cgenff_remas.rtf
   read rtf card unit 10 append
   close unit 10
   open unit 10 read card name @TOPPAR/par_all36_cgenff_remas.prm
   read para card flex unit 10 append
   close unit 10
   

The CGenFF paper is available in "Early View" from the JCC website:
http://www3.interscience.wiley.com/journal/122486907/abstract

This paper is mandatory reading for anyone who wants to do parameterization in the framework of CGenFF.

I now made a dedicated CGenFF download page. It gets updated more regularly than the CGenFF section on the MacKerell lab's force field page and also contains a bit more information about the CGenFF version you're downloading.

Also, CGenFF version 2b4 is out. It can be downloaded at the CGenFF download page (the CGenFF section on the MacKerell lab's force field page has not been updated yet at the time of this posting) and will also be included in the upcoming c36a4 developmental CHARMM version. Here is a cumulative list of changes since CGenFF version 2a5 (the version on which the CGenFF paper was based):

• Introduced carbamate parameters
  
• Introduced parameters for extended ketones (contributed by yapolyak following a discussion on this forum).
  
• Added limited thiocarbonyl support
  
• Added parameters for phenylhydrazine
  
• Added various other model compounds and parametes
  
• Updated bonded parameters for -CH2-NH2(+)-CH2 group (somewhat developmental; future tweaks likely)
  
• Corrected charges in RESI PPI2 topology
  
• Slight tweak in bonded parameters for benzenes with aliphatic substituents
  
• Slight tweak in benzamide bonded parameters
  
• Slight tweak in bonded parameters for sterane scaffold and beta-ionone ring
  
• Removed slack parameters containing NG2S0
  
• Fixed atom typing mistakes in cholesterol. Doing so mandated removing and adding a few parameters.
  
• Updated OCO angle in RESI DIOL.
  
• Corrected atom types on RESI GABD.

Lastly, the CGenFF paper is finally available in print. The citation is:
K. Vanommeslaeghe, E. Hatcher, C. Acharya, S. Kundu, S. Zhong, J. Shim, E. Darian, O. Guvench, P. Lopes, I. Vorobyov, A. D. MacKerell Jr., J. Comput. Chem. 2010, 31, 671-690.

In my lab, we are currently working with c35b1. Is it possible to use such force field or we need to update it to c36 (which is the c35b3).

Thank you

Please start a new thread for a new question rather than replying to an old one.

Also, please do some research before asking a question. Your answer is the very first item on the CGenFF FAQ:
http://dogmans.umaryland.edu/~kenno/cgenff/faq.html#compile
Yes, you can use CGenFF with c35, but you have to recompile with the CGENFF flag.

CGenFF version 2b5 can now be downloaded from the CGenFF download page as well as the CGenFF section on the MacKerell lab's force field page. Here is an incremental list of changes since CGenFF version 2b4 (copy-pasted from the download page):

• Changed atom typing scheme for phosphates; partial reparameterization of phosphor-oxygen bonds
  
• Eliminated asymmetric dihedrals and impropers - many compounds affected
  
• Big purge of "slack" parameters. If you are missing parameters that were present in older versions, please report it in this thread
  
• Renumbered atom types in the framework of CHARMM force field re-organization - old psf files are henceforward invalid
  
• Added carbamate parameters (RESI DMCA and DECA)
  
• Added parameters for styrene (RESI STYR)
  
• Miscellaneous new model compounds:
  • 2-Pyridone (RESI 2PYO) and uracil (RESI URAC)
    
  • 4,6-dioxabicyclo[3.3.0]octan-8-ol (RESI RSRF)
    
  • 1,2,4-Oxadiazole (RESI OXD4)
    
  • RESI HDZ1B, another phenylhydrazone model compound
    
  • methyl triphosphate (RESI METP)
• Updated parameters for 1,2,4-triazole (RESI TRZ4)
  
• Updated OG312 L-J parameters
  
• Updated CG2R61 NG311 bond; affected compounds: RESI FEOZ, FETZ
  
• Fixed inconsistencies in conjugated double bond parameters - many compounds affected
  
• Fixed an incorrect pyridine parameter - affects most pyridines
  
• Fixed aniline amide torsion lacking multiplicity 1 term
  
• Fixed a copy-paste error in a dihedral parameters in RESI DMP2
  
• Fixed atom types and charges on fluor-substituted THF derivatives (RESI T2FD, T2FU, TMFD, TMFU, THNP)
  
• Fixed charges on "carbocyclic" nucleotide derivatives (RESI CNGU, CSGU, CNTH, CSTH, CNT1, CNT2, CNT3, CNCY, CSCY)
  
• Fixed CG2DC? atom types on RESI TMCH (not conjugated)
  
• Fixed CG2DC? atom types on RESI PRAC, PRAL, HDZ2 (conjugated)
  
• Fixed atom types on RESI PMHA
  
• Fixed atom types on RESI 2AMP
  
• Fixed charges on RESI THF2
  
• Added missing improper in RESI 3CPY
  
• Fixed some IC tables
  
• Removed RESI 4HQI because it was broken while fixing parameters it shares with more important model compounds
  
• Removed RESI TEAZ, TDAZ because of issues with atom typing that were not trivial to correct - inquire if you need these compounds
  
• Removed RESI DMSO in preparation of upcoming sulfur reparameterization - inquire if you need it
  
• Renamed protonated TMAO to RESI TMAOP (neutral TMAO is coming soon; inquire if intersted)

Just in case someone missed Alex' announcement in the new sticky topic, the CGenFF program for automatic atom typing and assignment of parameters and charges by analogy is finally available! At the time this announcement is written, the available version of the CGenFF program is 0.9.0, which requires CGenFF version 2b6.

CGenFF 2b6 can now be downloaded from the CGenFF download page as well as the MacKerell lab's force field page. Here is an incremental list of changes since CGenFF version 2b5 (copy-pasted from the download page):

• Reorganization of improper dihedrals; impropers are now assigned using simple consistent rules and wildcards have been eliminated. Many parameters were tweaked and many compounds changed energy because of this.
• Renumbered atom types following the insertion of two new sulfur types (see below) - old psf files are henceforward invalid
• Added sulfonamide parameters
• Added sulfone parameters
• Added sulfoxide parameters
• Added anionic sulfonate parameters
• Added limited support for sulfonic esters
• Added parameters for ketones conjugated to double bonds and aromatic rings; slight tweak of existing ketone parameters
• Added aliphatic and aromatic amidinium parameters
• Added parameters for acyclic enol ethers
• Fixed more inconsistencies in conjugated double bond parameters - in particular, cyclic enol ethers should be better-behaved
• Fixed a few minor inconsistencies in torsional paramters for amides of aliphatic carboxylic acids (acetamide, propanamide, butyramide,...)
• Miscellaneous new model compounds:
  • p-cyanotoluene (RESI 4CYT)
  • trimethylamine N-oxide (RESI TMAO)
  • dimethylpropanamide (RESI DMPR)
  • pentamidine (RESI PNTM)
• Changed the names of the following residues so that they don't coincide with other CHARMM force fields anymore:
  • ALAD → AANM
  • GLYP → AAPM
  • DMPA → DMEP
  • CHL1 → CLOL
  • CHNS → CLNS
  • CHM1 → CLM1
• Rounded "3rd digit" charges in 3-fluoropyridine (RESI 3FLP) for improved portability
• Updated default nonbonded settings to fshift & vfswitch

This is very welcome news. Thank you for all of your efforts.

I do have a question about submitting molecules to the paramchem website. The instructions state that molecules must be in the mol2 format. This, from what I've been able to determine, is a proprietary format. Is SYBYL required? From this post and searches of the charmm documentation, I gather that (a) it is not but (b) CHARMM will not save in this format. Are these conclusions correct?

Assuming that they are, it appears that VMD may be able to save in this format. Can it do this successfully for purposes of submission to the paramchem website?

Thanks again!

Please start a new thread for a new question rather than replying to an old one! Can a site admin please move the previous post and this post into a new thread?

Thank you for your interest. The mol2 format was developed by the company Tripos (the producer of Sybyl), but they have included a full description of the format in their documentation for as long as I can remember, and have never discouraged other companies from using it as far as I know (just google "mol2 format" and the first link will let you download a pdf document with the full mol2 specification from the Tripos website). Together with the fact that they were one of the first actors in the field to develop a flexible molecular database format for drug discovery, this caused the mol2 format to become a de facto standard; mol2 is more widely supported than any other format and it is precisely for this reason that we chose it as our input format.

Be a scientist - try it! If VMD complies with the aforementioned mol2 standard, it should work (and these guys are pretty competent so I expect it will).