CCL: A review of force fields



 Sent to CCL by: "Fiona Case" [fhcase###hotmail.com]
 Hello,
 
I'd like to suggest that you have two questions to answer: Which forcefield to use, and which method to use (MD/MC, which flavor?). If you make a wise choice of forcefield, but a poor choice of method, you will still get poor results.
 
Also, when you look at many of the "forcefield review" papers they consider intramolecular interactions (bond lengths, bond angles, vibrational frequencies obtained from isolated molecules, etc.). Unless you are trying to predict properties of molecules in a vacuum, intermolecular interations are at least as important. Most do not consider the effect of temperature or pressure, nor do they provide guidence on the transferability of the forcefields (what will happen if your molecule is slightly different to the one used in the published study?)
 
This lack of validation/verification/guidence has been identified as a major limitation to the useful application of force field based methods.
 http://www.chemicalvision2020.org/pdfs/compchem.pdf
 http://fluidproperties.org/Vision-and-Strategic-Plan-5-Feb-06.pdf
 
However, there is some good news: Some organizations are attempting to solve this problem. If you take a look at the publications and websites that have resulted from their efforts you will already see some guidence on forcefield and method selection for "useful" simulations.
 
1) The Industrial Fluid Properties Simulation Collective (of which I am a member)http://fluidproperties.org/
 2) The BioSimGrid, http://www.biosimgrid.org/
 3) The protein folding challenges: http://predictioncenter.org/
 
4) If you are looking at something that is a crystal, see the Crystal Structure Prediction challenges: http://scripts.iucr.org/cgi-bin/paper?S0108768105016563
 
You could encourage every "modeling expert" you meet (particularly in academia) to support these efforts. The problems you are facing in trying to select a forcefield and method to use for your work are shared by everyone attempting to apply these methods. It is of benefit to us all that more validation/verification/guidence becomes available so that molecular simulations methods become "useful" to the wider world, and particularly in industry.
 Fiona Case
 ----Original Message Follows----
 
 From: "Carlos Nunez carlos_javierna*o*yahoo.com"
 <owner-chemistry#ccl.net>
 
 Reply-To: "CCL Subscribers" <chemistry#ccl.net>
 To: "Case, Fiona " <fhcase#hotmail.com>
 Subject: CCL: A review of force fields
 Date: Thu, 16 Mar 2006 20:57:41 -0500
 Sent to CCL by: "Carlos  Nunez" [carlos_javierna{:}yahoo.com]
 Hi,
 See this:
 MacKerell A. D. Jr. 2004. Empirical force fields for biological
 macromolecules: overview and issues. J. Comput. Chem. 25: 1584-1604.
 Hasta la proxima vez,
 Carlos N.http://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp-:-//www.ccl.net/chemistry/sub_unsub.shtmlhttp-:-//www.ccl.net/spammers.txt