From owner-chemistry@ccl.net Wed Jan 17 04:59:00 2007 From: "Steve Williams willsd%a%appstate.edu" To: CCL Subject: CCL: Gaussian QST2 Message-Id: <-33386-070116182537-25739-4/XmYvXzvr5FsXVTkJalKw|server.ccl.net> X-Original-From: Steve Williams Content-transfer-encoding: 7bit Content-type: text/plain; format=flowed; charset=ISO-8859-1 Date: Tue, 16 Jan 2007 17:08:02 -0500 MIME-version: 1.0 Sent to CCL by: Steve Williams [willsd:-:appstate.edu] Transition state optimizations are usually tricky, and always more difficult than equilibrium structure optimizations. I have had some success with taking a structure from a failed qst2 optimization (not always the last one, but one that seems "reasonable"), computing the frequencies to see if it is "close": only one imaginary frequency (or one that is significantly larger in magnitude than the others). If it is then this often works: # b3lyp gen opt=(ts,noeigentest,calcfc) (then whatever other keywords you need; you can also save time by reading in the force constants from previous frequency calculation if you have saved the checkpoint file) or in cases that still go wrong: # b3lyp gen opt=(ts,noeigentest,calcall) (this can be very slow, since an exact Hessian is used at each step) Good luck, Steve Williams Satyan Sharma satyan+/-sun3.oulu.fi wrote: > Sent to CCL by: Satyan Sharma [satyan:_:sun3.oulu.fi] > Hello everybody, > I am trying to do TS search using following route: > > #N B3LYP/GEN OPT=QST2 Nosymm TEST > > Optimization went fine for 8 steps. In the 8th step I see this: > Step number 8 out of a maximum of 510 > All quantities printed in internal units (Hartrees-Bohrs-Radians) > Update second derivatives using D2CorX and points 5 6 7 8 > Eigenvalues --- -0.00335 0.00229 0.00230 0.00232 0.00233 > Eigenvalues --- 0.00250 0.00293 0.00330 0.00377 0.00486 > Eigenvalues --- 0.00555 0.00584 0.00706 0.00930 0.00989 > > But in the 9th step : > Step number 9 out of a maximum of 510 > All quantities printed in internal units (Hartrees-Bohrs-Radians) > Update second derivatives using D2CorX and points 8 9 > Eigenvalues --- ********** 0.04188 0.04803 0.06493 0.06961 > Eigenvalues --- 0.08716 0.09281 0.11078 0.11260 0.13017 > > And the run stopped with following message: > 544 D251 0.00079 0.00079 -49.92858********** > 545 D252 0.00000 0.00000 -45.00426********** > 546 D253 0.00103 0.00103 -0.00008 0.00816 > RFO step: Lambda0=3.127208390D+13 Lambda=-8.54305189D-03. > Arithmetic Exception > > Please let me know if you have an idea. > Cheers, > Satyan Sharma> > > > From owner-chemistry@ccl.net Wed Jan 17 06:50:00 2007 From: "Guosheng Wu wu_guosheng2002 a yahoo.com" To: CCL Subject: CCL: Looking for outliers of protein-ligand scoring functions Message-Id: <-33387-070117064514-10946-6+E8RpCG4r9qfMCmEHcfpA{:}server.ccl.net> X-Original-From: "Guosheng Wu" Date: Wed, 17 Jan 2007 06:45:11 -0500 Sent to CCL by: "Guosheng Wu" [wu_guosheng2002],[yahoo.com] Dear colleagues, There have been many kinds of commericial software designed to predict the binding affinities for protein-ligand complexes, but unfortunately many times it's found out the errors can be very big. Personally I am quite interested in collecting the outliers from such kind of study: given a high resolution protein pdb file, and experimental binding affinity(Ki, or IC50 if Ki is not available) for a molecule, the predicted affinity(logKi) is more than 100 times(2 log units) from the experimental value, or from the equation one applied. Presumely, the published benchmark study should already have such kind of information, however for most of the time it's not true. One can see the results were simply given in a plot with lots of dots, or just a r^2 or q^2 values, but there is no much information about which structures have good predictions, or which are the obvious outliers. I would appreciate if you would like to share this information with me. Please let me know the following information: (1) pdb code, the pdb file you used for the target, the ligand 3D file you scored in mol format, which can be the docked structure if not in the original xray complex; (2)the experimental affinity, and the predicted value you got, or what kind of affinity that can be transfered from the units in the software. (3)(optional)The software you used and its version. As a reward, I promise I would analyze the information you send to me, and would send you my comments about possible issues in the structure files, or possible reasons why it's poorly scored. Please note all data should be in the public domain (pdb xrays, and disclosed 2D ligand structures). And the error should be more than 2 log units (or 100 times of binding affinity). Thanks a lot, Guosheng Wu From owner-chemistry@ccl.net Wed Jan 17 09:21:00 2007 From: "Adil Zhugralin zhugrali-*-bc.edu" To: CCL Subject: CCL: g03 density functionals for organometallic complexes Message-Id: <-33388-070116205643-2400-XTz9WzxO6kfx4/qX1Q42Zw[a]server.ccl.net> X-Original-From: "Adil Zhugralin" Date: Tue, 16 Jan 2007 20:56:40 -0500 Sent to CCL by: "Adil Zhugralin" [zhugrali-$-bc.edu] Dear all: I am trying to execute calculations on catalytic cycle involving ruthenium carbenes. I am experimentalist, but these calculations are essential for my research. Based on the little that I've read so far, DFT methods like B3LYP are generally not so good for organometallics. People in my field use either BP86 or PW91, but none of these are any good for calculating kinetic barriers. Recently Truhlar and co-workers disclosed a few new meta-hybrid-GGA's like M05 and M06, so I was wondering if anyone knows anything about them, that is to say, how reliable are they? Also, how would one go about obtaining them if necessary? And the last question: What basis sets would perform better for organometallics? (I'm trying to use 6-31G* on all atoms except Ru, for which I'm using Stuttgart/Dresden RECP). Thank you all! Adil From owner-chemistry@ccl.net Wed Jan 17 10:17:00 2007 From: "LI Daobing lidaobing*o*gmail.com" To: CCL Subject: CCL: LOG file size generated by TINKER Message-Id: <-33389-070117065856-14811-sqe9b9K9miWrUray3xE/yQ|server.ccl.net> X-Original-From: "LI Daobing" Content-Disposition: inline Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Wed, 17 Jan 2007 19:29:22 +0800 MIME-Version: 1.0 Sent to CCL by: "LI Daobing" [lidaobing^^^gmail.com] Hello, On 1/8/07, Chris Martin Wood c.m.wood:_:ljmu.ac.uk wrote: > > Sent to CCL by: "Chris Martin Wood" [c.m.wood(_)ljmu.ac.uk] > Dear Sir/Madam, > > Firstly, we would like to say what an excellent package TINKER is - thanks for making it freely available. > > We have used TINKER to examine the energy content of each atom in a small molecule. The data was generated by using ANALYZE, with the 'A' option. The program runs well and finishes. However, the LOG file generated by the program, and which contains the data we want, is 700 Mbytes in size. This size of file is inpossible for WORD or WORDPAD to read. The molecule is only small, and all atoms must be active. In other words, using the INACTIVE control word in a .KEY file is not an option. > > We were wondering if you could possibly help us, by explaining how we can get a number of smaller LOG files, such that we can open them and access the data. We are using the XP version of TINKER. > 1. you can use gvim or ultraedit, it should support larger size than word or wordpad 2. you can write your own program to analyze it instead of word and wordpad, 3. use grep or awk or something else to filter out unusable informations 4. split it into serveral little files, then you can read it with word or wordpad -- LI Daobing From owner-chemistry@ccl.net Wed Jan 17 10:52:01 2007 From: "jpp:_:lct.jussieu.fr" To: CCL Subject: CCL:G: overlays/gaussian 03 Message-Id: <-33390-070117095324-28350-E/2kdxRdWHyZwKi/UaeVRg-,-server.ccl.net> X-Original-From: jpp[*]lct.jussieu.fr Content-Transfer-Encoding: 8bit Content-Type: text/plain;charset=iso-8859-1 Date: Wed, 17 Jan 2007 14:44:18 +0100 (CET) MIME-Version: 1.0 Sent to CCL by: jpp_+_lct.jussieu.fr Dear all, I have a question about the G03 overlays or IOps. Is someone familiar with the "L" keywords? I managed to use the classical IOPs but have some trouble to use the ones related to a specific L number (as quoted in the manual). example : IOP(1/43) L116 : extent of REaction Field 0 dipole 1 quadrupole etc... Any ideas? Sincerely yours J-P Piquemal -- Dr. Jean-Philip Piquemal, Phd Maitre de Confιrences (Assist. Prof.), Laboratoire de Chimie Thιorique, Universitι Pierre et Marie Curie (Paris VI), | email : jpp=-=lct.jussieu.fr UMR 7616 du CNRS, Case Courrier 137, | tel. : (33) 1 44 27 25 04 4, place Jussieu, 75252 Paris cedex 05 | fax : (33) 1 44 27 41 17 Web : www.lct.jussieu.fr/pagesperso/jpp Tιl secrιtariat : (33) 1 44 27 38 79 From owner-chemistry@ccl.net Wed Jan 17 11:27:00 2007 From: "Masao Fiegl at5gomi.(a).rose.plala.or.jp" To: CCL Subject: CCL:G: A supercomputer center Message-Id: <-33391-070117094639-23924-lph1d/9M5cpjiJA+e9QFsw(a)server.ccl.net> X-Original-From: Masao Fiegl Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-2022-JP Date: Wed, 17 Jan 2007 15:47:25 +0100 MIME-Version: 1.0 Sent to CCL by: Masao Fiegl [at5gomi ~ rose.plala.or.jp] Dear CCL members, Thank you for many replies for my previous question. In addition, I am looking for a supercomputer center at North America and Europe where anyone can use Gaussian 03 at cheap charge. I am very grateful for any suggestions. Thanks for your help. Masao From owner-chemistry@ccl.net Wed Jan 17 12:19:00 2007 From: "Sidney Ramos Santana santanasidney(!)yahoo.com.br" To: CCL Subject: CCL:G: Strange Message in G03 output Message-Id: <-33392-070117115726-5096-e5ohgswvS9bgfK0Plpc8jw_+_server.ccl.net> X-Original-From: "Sidney Ramos Santana" Date: Wed, 17 Jan 2007 11:57:23 -0500 Sent to CCL by: "Sidney Ramos Santana" [santanasidney : yahoo.com.br] Dear CCL Users, I am trying to run a geometry optimization in the Gaussian03 in a PC with Linux CentOS 4.4, kernel 2.6.9, with 4GB of RAM memory and 100GB of free "scratch" partition . I am using ub3lyp/6-311++G** over 38 atoms and 566 basis functions. After a long time, it stops and writes the following messsage: "User's command returned 256" I would appreciate if anyone can tell me what means this message and what can I do to finish well this calculation? Thanks for any help! Best Regards, Sidney R. Santana santanasidney_at_yahoo.com.br From owner-chemistry@ccl.net Wed Jan 17 12:53:01 2007 From: "Marcel Swart marcel.swart{}icrea.es" To: CCL Subject: CCL: g03 density functionals for organometallic complexes Message-Id: <-33393-070117123148-20681-3m8AiSjCiSy584UCtvCixg---server.ccl.net> X-Original-From: Marcel Swart Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=WINDOWS-1252; delsp=yes; format=flowed Date: Wed, 17 Jan 2007 17:31:06 +0100 Mime-Version: 1.0 (Apple Message framework v752.3) Sent to CCL by: Marcel Swart [marcel.swart::icrea.es] Please don't forget the functionals containing optimized exchange (OPTX) like OPBE and OLYP; not only are they GGA's, thus much faster than hybrids, also they perform very well for spin states of organometallic systems, accuracy of geometries, as well as the energies and structures of transition states. In a recent study where we compared DFT with CCSD(T) calculations, the best performance was observed by mPBE0KCIS/QZ4P//OLYP/QZ4P. The same methodology has been used in a recent study on nucleophilic substitution at phosphorus. See for instance: M. Swart, A.W. Ehlers and K. Lammertsma "Performance of the OPBE exchange-correlation functional" Mol. Phys. 2004, 102, 2467-2474 M. Swart, A.R. Groenhof, A.W. Ehlers and K. Lammertsma "Validation of exchange-correlation functionals for spin states of iron-complexes" J. Phys. Chem. A 2004, 108, 5479-5483 M.A. van Bochove, M. Swart and F.M. Bickelhaupt "Nucleophilic substitution at phosphorus (SN2/./P): Disappearance and reappearance of reaction barriers" J. Am. Chem. Soc. 2006, 128, 10738-10744 M. Swart, M. Solΰ and F.M. Bickelhaupt "Energy landscapes of nucleophilic substitution (SN2) reactions: A comparison of density functional theory and coupled cluster methods" J. Comput. Chem. 2007, accepted for publication On Jan 17, 2007, at 2:56 AM, Adil Zhugralin zhugrali-*-bc.edu wrote: > > Dear all: > > I am trying to execute calculations on catalytic cycle involving > ruthenium carbenes. I am experimentalist, but these calculations > are essential for my research. Based on the little that I've read > so far, DFT methods like B3LYP are generally not so good for > organometallics. People in my field use either BP86 or PW91, but > none of these are any good for calculating kinetic barriers. > Recently Truhlar and co-workers disclosed a few new meta-hybrid- > GGA's like M05 and M06, so I was wondering if anyone knows anything > about them, that is to say, how reliable are they? Also, how would > one go about obtaining them if necessary? And the last question: > What basis sets would perform better for organometallics? (I'm > trying to use 6-31G* on all atoms except Ru, for which I'm using > Stuttgart/Dresden RECP). > > Thank you all! > > Adil –––––––––––––––––––––––––––––––––––––––––––– dr. Marcel Swart ICREA researcher at Institut de Quνmica Computacional Universitat de Girona Campus Montilivi 17071 Girona Catalunya (Spain) fax +34-972-418356 e-mail marcel.swart/./icrea.es marcel.swart/./udg.es web http://www.icrea.es/pag.asp?id=Marcel.Swart http://iqc.udg.es/~marcel –––––––––––––––––––––––––––––––––––––––––––– From owner-chemistry@ccl.net Wed Jan 17 15:55:00 2007 From: "Wai-To Chan chan[a]curl.gkcl.yorku.ca" To: CCL Subject: CCL:G: g03 density functionals for organometallic complexes Message-Id: <-33394-070117154216-8702-XQgbHFkI4CAcusvWbDmiOg]^[server.ccl.net> X-Original-From: Wai-To Chan Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii Date: Wed, 17 Jan 2007 15:18:06 -0500 (EST) MIME-Version: 1.0 Sent to CCL by: Wai-To Chan [chan===curl.gkcl.yorku.ca] <<<<<<<<<<<<<<<<<<<<<< I am trying to execute calculations on catalytic cycle involving ruthenium carbenes. I am experimentalist, but these calculations are essential for my research. Based on the little that I've read so far, DFT methods like B3LYP are generally not so good for organometallics. People in my field use either BP86 or PW91, but none of these are any good for calculating kinetic barriers. Recently Truhlar and co-workers disclosed a few new meta-hybrid-GGA's like M05 and M06, so I was wondering if anyone knows anything about them, that is to say, how reliable are they? Also, how would one go about obtaining them if necessary? And the last question: What basis sets would perform better for organometallics? (I'm trying to use 6-31G* on all atoms except Ru, for which I'm using Stuttgart/Dresden RECP). Thank you all! Adil >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>. The SDD ECP is a reliable basis set for transition metals. By default it is the D95 basis set which is assigned to the non-metallic element if you specify SDD as the basis set. For more reliable results I would use D95* instead of D95. With G98 (not sure about G03) I would specify in the input card something like 'B3LYP/SDD extrabasis' You have to include the polarization function data in the input file for the atoms other than Ru. You can obtain the basis set details using the keyword 'gfprint' Wai-To Chan From owner-chemistry@ccl.net Wed Jan 17 18:32:00 2007 From: "Xie Wang xie.wang++gmail.com" To: CCL Subject: CCL:G: casscf dipole/transition moment Message-Id: <-33395-070117183004-8304-vuw0iiSFXXZWZYxzEDJfdg{:}server.ccl.net> X-Original-From: "Xie Wang" Date: Wed, 17 Jan 2007 18:30:01 -0500 Sent to CCL by: "Xie Wang" [xie.wang,,gmail.com] Hello, I am using g03 casscf. Does anyone know how to get ground and excited states dipole moment and transition moment between ground and excited states? Any help is appreciated! Thanks! Xie From owner-chemistry@ccl.net Wed Jan 17 19:07:01 2007 From: "Edvin Erdtman edvin.erdtman()nat.oru.se" To: CCL Subject: CCL: Gromacs: sp2-carbon Message-Id: <-33396-070117101409-5830-GfsV69P2lpZpHQvNc3AtUg]^[server.ccl.net> X-Original-From: "Edvin Erdtman" Date: Wed, 17 Jan 2007 10:14:05 -0500 Sent to CCL by: "Edvin Erdtman" [edvin.erdtman~!~nat.oru.se] Hello I am trying to run a calculation of a membrane with a molecule inside it, containing a double bond. I tried to use PRODRG, but it generates a strange topology file. The representations for sp2 carbons seems to be incorrect... The carbons in the double bond get C and CR61 representations respectively. The bond length from CR61 to next (sp3) carbon become 0.139 nm, (as another double bond). When I am running the simulation, there is too much movements around the double bond, so there must be something strange here... I tried to change the representation of the "C" carbon to "CB" or "CR6" but it doesn't seem to help... Thankful for your help! -- Edvin Erdtman Ph.D. student rebro University edvin.erdtman-.-nat.oru.se From owner-chemistry@ccl.net Wed Jan 17 19:42:00 2007 From: "Marek Sierka marek.sierka++chemie.hu-berlin.de" To: CCL Subject: CCL: Turbomole support forum at www.turbo-forum.com Message-Id: <-33397-070117184546-14634-/Ez8fiCFroF3y6zKhq+QPA]-[server.ccl.net> X-Original-From: "Marek Sierka" Date: Wed, 17 Jan 2007 18:45:42 -0500 Sent to CCL by: "Marek Sierka" [marek.sierka!^!chemie.hu-berlin.de] Dear CCL subscribers, I would like to announce a freshly started support forum for users of the quantum chemical package Turbomole at www.turbo-forum.com. The purpose of this forum is to provide an online interaction platform for Turbomole users and developers. We all encounter different problems and issues when running the program for our everyday calculations and the idea is to share our experience with others. The problems and questions posted in this forum will also provide feedback and inspiration for code developers. You are welcome to post all your questions and remarks as well as share your experience with others. Best Wishes Marek Sierka -- -------------------------------------------------------------- Dr. Marek Sierka Tel. +(49) (30) 20937477 Theoretical Chemistry FAX +(49) (30) 20937136 Humboldt University, Unter den Linden 6 D-10099 Berlin E-mail marek.sierka(-)chemie.hu-berlin.de --------------------------------------------------------------