From chemistry-request@server.ccl.net Fri Aug 17 07:10:01 2001 Received: from earth.ox.ac.uk (root@[163.1.22.6]) by server.ccl.net (8.11.0/8.11.0) with ESMTP id f7HBA0v01654 for ; Fri, 17 Aug 2001 07:10:00 -0400 Received: from pauling.earth.ox.ac.uk (pauling [163.1.22.22]) by earth.ox.ac.uk (8.9.3+Sun/8.9.1) with ESMTP id MAA23351 for ; Fri, 17 Aug 2001 12:09:58 +0100 (BST) From: Keith Refson MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15228.64389.348325.335076@gargle.gargle.HOWL> Date: Fri, 17 Aug 2001 12:09:57 +0100 To: Computational Chemistry List Subject: SUMMARY: Potential parameter fitting X-Mailer: VM 6.95-rfhacked under 21.1 (patch 12) "Channel Islands" XEmacs Lucid I received a few attempts to summarise the responses, so here is a brief summary. I asked what programs were avialable for potential parameter fitting to energy hypersurfaces, such as ab-initio energies. The main responses were 1. NBP by the late Prof D. E. Williams, http://www.louisville.edu/~dewill01/nbp.html This seems to be about the best suited program for such tasks. It used to be obtainable directly from Prof Williams, but none of my respondents had any idea how to obtain it now, as he recently passed away. I hope this does not become lost to the community. 2. A number of respondents pointed me to GULP by Julian Gale http://www.ch.ic.ac.uk/gale/Research/gulp.html which is slightly embarrassing since I thought I knew that program fairly well :) It can indeed be used to fit to an energy hypersurface (which I erroneously thought it wasn't well suited for). GULP uses a 3D periodic model, which isn't what I need for this particular calculation, (A simple isolated moecular cluster), but of course it's not to hard to achieve this by judicious use of a large supercell and interaction cutoffs and turning off the long-ranged part of coulomb interactions. 3. Though not what I was looking for Andrew Horsfield (horsfield@fecit.co.uk)suggested that The OXON tight binding package also fits (and implements) Tersoff type potentials. 4. "Kjeld Rasmussen studied with Lifson and has modified that august program over the years to be a parameter-fitting tool. It can fit to structural parameters and vibrational frequencies but i do not believe to a hypersurface. Kjeld readily gives the program away. Some of the background to his program is given at: [1]http://struktur.kemi.dtu.dk/cff/cffhome.html" 5. "Arnold Hagler had a similar program (I believe called "Probe") that also came from Lifson's program and it was modified to fit hypersurface data but I think it is now a proprietary commercial product for MSI consortium members." 6. Marcelo Zaldini has been "developing a methodology, called SJBR, to do the parametrization of the intermolecular terms (coulomb + lennard_jones), based upon ab initio calculations, in an automatic fashion." which he hopes to release at some future time, when it is finished. 7. Other co-respondents recommended general least squares fitting methods such as the least-squares module in Scientific Python, available from http://dirac.cnrs-orleans.fr/programs/scientific.html, or even gnuplot! Although of course the built-in fit is only 1-dimensional and therefore not suitable, the gnuplot FAQ contains links to a number of other general purpose leasy-squares programs, such as Fudgit. None of these really fulfil the spirit of my request which was for a program which handles the bookkeeping and set-up for molecular models. Thanks to Konrad Hinsen , Manuel Melle-Franco , Irena Efremenko , Andrew Rohl , Brian J. Teppen" , Marcelo Zaldini Hernandes , Daquan Gao Andrew Horsfield for the suggestions. Keith Refson -- Dr Keith Refson, "Paradigm is a word too often used by those who would Dept of Earth Sciences like to have a new idea but cannot think of one." Parks Road, -- Mervyn King, Deputy Governor, Bank of England Oxford OX1 3PR, UK Keith.Refson@ Tel: 01865 272026 earth.ox.ac.uk Fax: 01865 272072