From peha@sgiclu.chemie.uni-konstanz.de Sat Feb 28 06:14:48 1998 Received: from sg7.chemie.uni-konstanz.de for peha@sgiclu.chemie.uni-konstanz.de by www.ccl.net (8.8.3/950822.1) id FAA06993; Sat, 28 Feb 1998 05:54:33 -0500 (EST) Received: from sgiclu.chemie.uni-konstanz.de (sg17.chemie.uni-konstanz.de [134.34.7.91]) by sg7.chemie.uni-konstanz.de (8.8.6/8.8.6) with SMTP id LAA24733 for <@sg7.chemie.uni-konstanz.de:chemistry@www.ccl.net>; Sat, 28 Feb 1998 11:55:26 +0100 (MET) Received: by sgiclu.chemie.uni-konstanz.de (950413.SGI.8.6.12/940406.SGI) for chemistry@www.ccl.net id LAA07152; Sat, 28 Feb 1998 11:55:16 +0100 From: peha@sg17.chemie.uni-konstanz.de (Peter Happersberger) Message-Id: <9802281155.ZM7150@sg17.chemie.uni-konstanz.de> Date: Sat, 28 Feb 1998 11:55:11 +0100 X-Mailer: Z-Mail (3.2.3 08feb96 MediaMail) To: chemistry@www.ccl.net Subject: MD simulation of proteins with TINKER Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Dear CClers I would like to do some MD simulations of solvated proteins (e.g. BPTI) with the program package TINKER V3.5 (http://dasher.wustl.edu/tinker/) under Linux. But till now, I couldn't generate a TINKER specific start file (*.xyz). What I have done: 1. I have generated a solvent volume around the protein BPTI (pdb1bpi.ent) with the program SOLVATE (http://www.imo.physik.uni-muenchen.de/~grubi/solvate/docu.html). 2. After conversion of the pdb-file to the tinker specific xyz-file (TINKER program 'pdbxyz') I have started the MD simulation with the TINKER program 'dynamic'. The program stopped after a short time with several error messages, such like Enter Potential Parameter File Name : amber Undefined or Illegal Atom Types : Type Atom Number Atom Type Atom 1 0 Atom ....... ............ Atom 446 0 Atom 447 0 Atoms with an Unusual Number of Attached Atoms : Type Atom Name Atom Type Expected Found Valence 1-OH2 0 0 2 Valence 2-H1 0 0 1 ................... Valence 447-H2 0 0 1 MECHANIC -- Some Required Potential Energy Parameters are Undefined TINKER is Terminating due to the Occurrence of a Severe Error By contrast, the MD simulation without the water shell (e.g. only the protein) succeded. I am appreciated greatly if any one give me some information how I can generate a "correct" xyz-file. Could anybody send me a "correct" example files (xyz-, seq-, pdb-, key-, ...) which are neccessary to start a MD simulation? Thank you, Peter Happersberger -- ------------------------------------------------------------------------ Peter Happersberger Ph.D. Student Faculty of Chemistry University of Konstanz Email: Peter.Happersberger@uni-konstanz.de URL: http://sg17.chemie.uni-konstanz.de/~peha/ Voicenet: ++49-(0)7531-88-3928 or Fax: ++49-(0)7531-88-3097 Papernet: AG Przybylski, P.O. Box M 732, 78457 Konstanz, Germany ------------------------------------------------------------------------- From elewars@alchemy.chem.utoronto.ca Sat Feb 28 17:14:54 1998 Received: from alchemy.chem.utoronto.ca for elewars@alchemy.chem.utoronto.ca by www.ccl.net (8.8.3/950822.1) id QAA08445; Sat, 28 Feb 1998 16:19:07 -0500 (EST) Received: (from elewars@localhost) by alchemy.chem.utoronto.ca (8.7.4/8.7.3) id QAA21928 for chemistry@www.ccl.net; Sat, 28 Feb 1998 16:19:07 -0500 (EST) Date: Sat, 28 Feb 1998 16:19:07 -0500 (EST) From: "E. Lewars" Message-Id: <199802282119.QAA21928@alchemy.chem.utoronto.ca> To: chemistry@www.ccl.net Subject: SUGGESTION ABOUT TS OPTIMIZATION 1998 Feb 28 Ken Legendre wrote: 27-FEB-1998 18:45:26.52 >From: IN%"kwlegen@rs6000.cmp.ilstu.edu" "Ken Legendre" To: IN%"chemistry@www.ccl.net" CC: Subj: CCL:G:Transition State calculations Dear CCLers Hello. I am trying to calculate transition states, using the tight option in Gaussian 94, at very high levels of theory, and having problems. I have tried the TS command and get a wrong number of eigenvlaues error. I have also tried the QST2 and QST3 and get results but the calculation does not converge. I also used the maxcycle command in order to give the QST2 and QST3 calculations more cycles to converge. The program quits after 100 cycles, even when the number of cycles is set to larger number of cycles. Im not sure why at this point it is doing that. I have also tried the calcall command and also get the wrong number of eigenvalues. I would appreciate any suggestions. Thanks in advance Kenneth Legendre kwlegen@rs6000.cmp.ilstu.edu Illinois State University Department of Chemistry Normal, Il 61761 ============== 1) Use the keyword NoEigenTest. In my experience Gaussian almost never finds the TS if it has to check the number of negative eigenvalues of the Hessian. When you get a putative TS, you can check it out in a separate frequency job for the number of Hessian eigenvalues (the number of imaginary freqs in a freq calc). 2) Find the TS at a lower level of theory and use that structure as input, requesting a HF calculation to get an initial Hessian. 3) A TS that exists at one level of theory may not at another level. ========== E. Lewars ==========