From jerzy@fqs.fujitsu.co.jp Mon Oct 31 02:22:26 1994 Received: from fgwmail.fujitsu.co.jp for jerzy@fqs.fujitsu.co.jp by www.ccl.net (8.6.9/930601.1506) id CAA24336; Mon, 31 Oct 1994 02:18:54 -0500 Received: from fdm.fujitsu.co.jp by fgwmail.fujitsu.co.jp (8.6.9+2.4W/3.3W5-MX941027Fujitsu Mail Gateway) id QAA19647; Mon, 31 Oct 1994 16:18:46 +0900 Received: from celica by fdm.fujitsu.co.jp (5.65/6.4J.6) id AA17520; Mon, 31 Oct 94 16:09:57 +0900 Received: from localhost (jerzy@localhost) by fqs.fujitsu.co.jp (8.6.4/3.3Wb-) id PAA27456 for chemistry@ccl.net; Mon, 31 Oct 1994 15:08:00 +0900 Date: Mon, 31 Oct 1994 15:08:00 +0900 From: Jerzy Rudzinski Message-Id: <199410310608.PAA27456@fqs.fujitsu.co.jp> To: chemistry@ccl.net Subject: Summary - quadrapole moments ORIGINAL MESSAGE: > > Dear Friends, > > I would very much appreciate receiving pointers to the public domain or > commerial software that can calculate (after optization of the structure) > quadropole moments of sizable molecules (about 100 atoms). > Thank you! With the best wish, Jerzy Rudzinski > =========================================================================== >>From: peon@medchem.dfh.dk (Per-Ola Norrby) >>Subject: Re: CCL:quadrapole moments Jerzy, This can be done by Gaussian92, using a semi-empirical Hamiltonian (eg, AM1). It's far from public domain, academic price is approx. $2000. It runs on any number of workstations and supercomputers. More information can be obtained by mailing info@gaussian.com Per-Ola Norrby * Per-Ola Norrby * The Royal Danish School of Pharmacy, Dept. of Med. Chem. * Universitetsparken 2, DK 2100 Copenhagen, Denmark * tel. +45-35376777-506, +45-35370850 fax +45-35372209 * Internet: peon@medchem.dfh.dk =========================================================================== >>From: salahub@ERE.UMontreal.CA (Salahub Dennis) >>Subject: CCL:quadrapole moments (fwd) Density Functional Methods should be able to handle this problem, although it would be demanding. Since I see you are at Fujitsu, I might assume that you can find substantial computer power... The optimization step would be the most costly. IF it were possible to optimize with molecular mechanics or a semiempirical scheme, then fixed geometry DFT would be far less costly. You could try any one of several programs. DGAUSS (from Cray...), deMon (our program), DEFT (Alain St-Amant at the University of Ottawa) for the Gaussian family. (Also GAUSSIAN, CADPAC etc. but I don't think they can do 100 atoms, because the coulomb part is done exactly, with all the integrals, whereas DGAUSS etc. use fits to auxiliary basis functions), DMOL (from Biosym) uses numerical atomic orbitals; the folks at Biosym MIGHT(?) like the challenge??..., ADF from Evert-Jan Baerends' group in Amsterdam can use Slater orbitals and also numerical or gaussian functions. It might be worth checking some of these out on smaller molecules related to those that interest you, before tackling the 100-atom cases. Good luck! -Dennis Salahub -- ************************************************************ * Dennis Salahub SALAHUB@ERE.UMONTREAL.CA * * Departement de chimie Tel. (514) 343-6755 * * Universite de Montreal Fax (514) 343-2468 * * C.P. 6128, Succ. Centre-ville * * Montreal, Quebec H3C 3J7 * * Canada * ************************************************************ =========================================================================== >>From: monge@anny.psgvb.com (Alessandro Monge) >>Subject: Re: CCL:quadrapole moments Dear Jerzy, Schrodinger's ab initio quantum chemistry code, PS-GVB, allows the calculation of multipole moments, from dipole through hexadecapole. Geometry optimization can be carried out at the HF and HF/GVB level with substantial CPU time savings as compared to any existing commercial quantum chemistry program. Timing advantages are in the order of 5-10 with respect to Gaussian92 for large molecules, depending upon basis set and machine. PS-GVB uses novel numerical technology (based on Pseudo-Spectral methods, originally developed for hydrodynamic simulations of turbolence) involving both basis sets and numerical grids to solve the equations of electronic structure theory. PS-GVB is capable of performing calculations involving up to 200 atoms, 5000 basis functions, 50 GVB pairs for GVB calculations and is available on a variety of platforms. Alessandro Monge ------------------------------------------------------------------------ Alessandro Monge Schrodinger, Inc. monge@psgvb.com (800) 20-PSGVB ========================================================================= >>From: Wolfgang Sauer >>Subject: Re: CCL:quadrapole moments Dear Jerzy Rudzinski, our semiempirical program package VAMP will give you electrical multipoles up to currently octupoles. They are usually used e.g. to fit NAO point charges to give the best representation of a molecule's ESP, or as property descriptors in QSAR studies with our program Tsar through a transparent interface. The only limit for the size of your molecular system lies in your hardware; I myself use Vamp routinely on small peptides of about 100 heavy atoms (plus ca. 100 hydrogens) on modern workstations without problem. If you would like to get more information on Vamp or have some other questions, please do not hesitate to contact either me or our distributor in Japan. All the best, Wolfgang Sauer. -- Wolfgang Sauer ------------ | ## ## ## | Oxford Molecular - CCC | ## ## ## | Naegelsbachstr. 25 | ###### | D-91052 ERLANGEN, FRG | ## | | ## | Tel : +49 9131 8796 0 | #### | Fax : +49 9131 8796 11 ------------ =========================================================================== From CHUCK@psipsy.uct.ac.za Mon Oct 31 04:10:19 1994 Received: from ucthpx.uct.ac.za for CHUCK@psipsy.uct.ac.za by www.ccl.net (8.6.9/930601.1506) id DAA24887; Mon, 31 Oct 1994 03:22:19 -0500 Received: by ucthpx.uct.ac.za (/\==/\ Smail3.1.28.1 #28.43) id m0r1rz3-000KFsC; Mon, 31 Oct 94 10:20 SAST Received: From ITS/WORKQ1 by charongw1.uct.ac.za via Charon-4.0A-VROOM with IPX id 100.941031101419.320; 31 Oct 94 10:22:31 -0200 Message-ID: To: sling@euclid.chem.washington.edu (Song Ling), chemistry@ccl.net From: "Marais, Charles F. " Date: Mon, 31 Oct 1994 10:20:22 SAST-2 Subject: Re: CCL:100% parallel MD Priority: normal X-mailer: WinPMail v1.0 (R1) > Will the > -dt, dt approach generate the "same" trajectory as the 2dt one if the > system is sensitive to initial conditions? This is very interesting. Has much work been done on chaos in MD ? Any pointers would be much appreciated. C Dr Charles Marais Department of Chemistry University of Cape Town Private Bag Rondebosch chuck@uctvax.uct.ac.za South Africa 7700 chuck@psipsy.uct.ac.za ------------------------------------------------------------------ From Bernd.Hartke@RUS.Uni-Stuttgart.DE Mon Oct 31 04:22:26 1994 Received: from dfunms.rus.uni-stuttgart.de for Bernd.Hartke@RUS.Uni-Stuttgart.DE by www.ccl.net (8.6.9/930601.1506) id EAA25172; Mon, 31 Oct 1994 04:12:34 -0500 From: Received: from servus02.rus.uni-stuttgart.de by dfunms.rus.uni-stuttgart.de with SMTP id AA22307 (5.65c8/DFUE-M1.0 for ); Mon, 31 Oct 1994 10:12:23 +0100 Received: from servus07.rus.uni-stuttgart.de by servus02.rus.uni-stuttgart.de (AIX 3.2/UCB 5.64/SERVus-1.2) id AA57242; Mon, 31 Oct 1994 10:12:22 +0100 Received: by servus07.rus.uni-stuttgart.de (AIX 3.2/UCB 5.64/4.03) id AA25615; Mon, 31 Oct 1994 10:12:22 +0100 Message-Id: <9410310912.AA25615@servus07.rus.uni-stuttgart.de> Subject: looking for graphics software To: chemistry@ccl.net Date: Mon, 31 Oct 1994 10:12:21 +0100 (MEZ) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1733 Dear CCL readers, our group of ab initio quantum chemists is looking for a suitable graphics software that fits our particular needs: We want - good 2D contour plots (including a really working labeling mechanism for contour lines), - 3D renderings of potential energy surfaces (perspective, hidden line, etc., but no fancy stuff), - perhaps also 3D isosurfaces, but this is not a must, - simple ball-and-stick plots of molecules and clusters (or the possibility to create these by a few simple steps), - of course simple x,y-graphs etc., - publication-quality laser printer output (postscript) All this should be realized as a subroutine library directly callable from our own Fortran programs. And it need not be public domain; we are actually willing to pay for it. We do NOT need/want (and hence do not want to pay for): - a black-box system, - an elaborate graphical user interface, - complicated, high-power graphics tools like photographic or video image processing, etc. For example, to name a few real-life products, a full-blown AVS or IDL package simply contains too many beautiful features we will never use; something like CA-DISSPLA would be more in our line; while GNUPLOT has a few shortcomings. We would like the graphics to run on HP735, Sun SPARC, and possibly Silicon Graphics machines. Any recommendations, hints, and your own positive or negative experiences are highly welcome! Bernd Hartke -- Dr. Bernd Hartke e-mail: bernd.hartke@rus.uni-stuttgart.de Dep. of Theoretical Chemistry e-mail: bxh@wiffin.chem.ucla.edu University of Stuttgart Pfaffenwaldring 55 Phone: +49-711-685-4409 70569 Stuttgart FAX: +49-711-685-4442 GERMANY From winn@tph12.tuwien.ac.at Mon Oct 31 05:22:27 1994 Received: from tph12.tuwien.ac.at for winn@tph12.tuwien.ac.at by www.ccl.net (8.6.9/930601.1506) id EAA25574; Mon, 31 Oct 1994 04:49:38 -0500 Message-Id: <199410310949.EAA25574@www.ccl.net> Received: by tph12.tuwien.ac.at (1.37.109.4/16.2) id AA17053; Mon, 31 Oct 94 10:52:27 +0100 From: Martyn Winn Subject: CCL:100% parallel MD To: chemistry@ccl.net Date: Mon, 31 Oct 94 10:52:27 MEZ Mailer: Elm [revision: 70.85] Two points: 1) My impression is that there is much work devoted to chaotic trajectories for small molecules (i.e. with a small no. of degrees of freedom) but that the issue is ignored for MD simulations of bulk systems. This is, of course, purely for reasons of convenience. 2) If one is in a chaotic regime, then I'm sure that the -dt, dt trajectory will differ from the 2dt trajectory, but taken individually each would be a perfectly valid trajectory. There would be no reason to prefer one over the other. In fact, a chaotic trajectory is surely ideal to satisfy the ergodic requirements of most MD simulations. I wonder if anyone is answering the original question about efficiency?! Martyn Winn -- ***************** Dr. Martyn Winn ************************************** \ / \ Institut fuer Theoretische Physik, Technische Universitaet Wien, / \ Wiedner Hauptstrasse 8-10, A-1040 Wien, Austria. / \____ / \ Tel: +43 1 58801 5678 / \ Fax: +43 1 5867760 / \ E-mail: winn@tph12.tuwien.ac.at / \ URL: http://tph.tuwien.ac.at/~winn/me.html / -------------------------------------------------- From CHMIGORN@NUSVM.NUS.SG Mon Oct 31 06:22:28 1994 Received: from phem3 for CHMIGORN@NUSVM.NUS.SG by www.ccl.net (8.6.9/930601.1506) id FAA25815; Mon, 31 Oct 1994 05:29:05 -0500 Received: from NUSVM.NUS.SG (MAILER@NUSVM) by phem3.acs.ohio-state.edu (PMDF V4.2-13 #5888) id <01HIX0A85NWW8YC14J@phem3.acs.ohio-state.edu>; Mon, 31 Oct 1994 06:28:58 EDT Received: from NUSVM.NUS.SG (NJE origin CHMIGORN@NUSVM) by NUSVM.NUS.SG (LMail V1.2a/1.8a) with RFC822 id 2557; Mon, 31 Oct 1994 18:31:10 +0800 Date: Mon, 31 Oct 1994 18:25:01 +0700 (SST) From: "I.Novak" Subject: VIEWERS FOR GAUSSIAN To: CHEMISTRY@ccl.net Message-id: <01HIX0A85NWY8YC14J@phem3.acs.ohio-state.edu> Content-transfer-encoding: 7BIT Dear Netters, here is the summary of responses received so far: a) XMol available from anonymous ftp.msc.edu pub/xmol Runs on Sun. Contact D.C.Doherty doherty@msc.edu b) Sybyl 6.1 supports G92 and runs on Sun c) AMPAC GUI will manipulate G92 files and runs on Sun contact: Semichem Inc. Dr A.Holder aholder@CCTR.UMKC.EDU Thanks to all who answered! I.Novak chmigorn@nusvm From dimitris@3dp.com Mon Oct 31 10:23:55 1994 Received: from boris.3dp.com for dimitris@3dp.com by www.ccl.net (8.6.9/930601.1506) id KAA29450; Mon, 31 Oct 1994 10:16:32 -0500 Received: from europa.3dp.com by boris.3dp.com via SMTP (931110.SGI/930416.SGI) for chemistry@ccl.net id AA23736; Fri, 28 Oct 94 17:13:39 -0400 Received: by europa.3dp.com (931110.SGI) id AA10480; Fri, 28 Oct 94 17:13:39 -0400 From: "Dimitris Agrafiotis" Message-Id: <9410281713.ZM10478@europa.3dp.com> Date: Fri, 28 Oct 1994 17:13:38 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: chemistry@ccl.net Subject: (Fwd) Returned mail: User unknown Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Does anyone know Peter Willett's e-mail address at the University of Sheffield in the UK? Thanks for your help. -- Dimitris K. Agrafiotis, PhD | e-mail: dimitris@3dp.com 3-Dimensional Pharmaceuticals, Inc. | tel: (215) 222-8950 3700 Market Street | fax: (215) 222-8960 Philadelphia, PA 19104 From dimitris@3dp.com Mon Oct 31 10:28:15 1994 Received: from boris.3dp.com for dimitris@3dp.com by www.ccl.net (8.6.9/930601.1506) id KAA29446; Mon, 31 Oct 1994 10:16:26 -0500 Received: from europa.3dp.com by boris.3dp.com via SMTP (931110.SGI/930416.SGI) for chemistry@ccl.net id AA23853; Fri, 28 Oct 94 17:23:29 -0400 Received: by europa.3dp.com (931110.SGI) id AA10480; Fri, 28 Oct 94 17:13:39 -0400 From: "Dimitris Agrafiotis" Message-Id: <9410281713.ZM10478@europa.3dp.com> Date: Fri, 28 Oct 1994 17:13:38 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: chemistry@ccl.net Subject: (Fwd) Returned mail: User unknown Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Does anyone know Peter Willett's e-mail address at the University of Sheffield in the UK? Thanks for your help. -- Dimitris K. Agrafiotis, PhD | e-mail: dimitris@3dp.com 3-Dimensional Pharmaceuticals, Inc. | tel: (215) 222-8950 3700 Market Street | fax: (215) 222-8960 Philadelphia, PA 19104 From dimitris@3dp.com Mon Oct 31 10:29:54 1994 Received: from boris.3dp.com for dimitris@3dp.com by www.ccl.net (8.6.9/930601.1506) id KAA29472; Mon, 31 Oct 1994 10:16:47 -0500 Received: from europa.3dp.com by boris.3dp.com via SMTP (931110.SGI/930416.SGI) for chemistry@ccl.net id AA23755; Fri, 28 Oct 94 17:16:09 -0400 Received: by europa.3dp.com (931110.SGI) id AA10487; Fri, 28 Oct 94 17:15:06 -0400 From: "Dimitris Agrafiotis" Message-Id: <9410281715.ZM10485@europa.3dp.com> Date: Fri, 28 Oct 1994 17:15:06 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: chemistry@ccl.net Subject: Looking for Peter Willett's e-mail address Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Does anyone know Peter Willett's e-mail address at the University of Sheffield in the UK? Thanks for your help. -- Dimitris K. Agrafiotis, PhD | e-mail: dimitris@3dp.com 3-Dimensional Pharmaceuticals, Inc. | tel: (215) 222-8950 3700 Market Street | fax: (215) 222-8960 Philadelphia, PA 19104 From dimitris@3dp.com Mon Oct 31 10:31:34 1994 Received: from boris.3dp.com for dimitris@3dp.com by www.ccl.net (8.6.9/930601.1506) id KAA29478; Mon, 31 Oct 1994 10:16:56 -0500 Received: from europa.3dp.com by boris.3dp.com via SMTP (931110.SGI/930416.SGI) for chemistry@ccl.net id AA23749; Fri, 28 Oct 94 17:15:28 -0400 Received: by europa.3dp.com (931110.SGI) id AA10484; Fri, 28 Oct 94 17:14:23 -0400 From: "Dimitris Agrafiotis" Message-Id: <9410281714.ZM10482@europa.3dp.com> Date: Fri, 28 Oct 1994 17:14:23 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: chemistry@ccl.net Subject: Need C/C++ libraries for clustering Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I am looking for public/commercial C/C++ libraries for clustering large amounts of multidimensional data. Your help would be greatly appreciated. I will summarize and submit the responses to the list. Thanks for your help. -- Dimitris K. Agrafiotis, PhD | e-mail: dimitris@3dp.com 3-Dimensional Pharmaceuticals, Inc. | tel: (215) 222-8950 3700 Market Street | fax: (215) 222-8960 Philadelphia, PA 19104 From KEWISE@aardvark.ucs.uoknor.edu Mon Oct 31 10:33:19 1994 Received: from aardvark.ucs.uoknor.edu for KEWISE@aardvark.ucs.uoknor.edu by www.ccl.net (8.6.9/930601.1506) id JAA28673; Mon, 31 Oct 1994 09:23:45 -0500 From: Message-Id: <199410311423.JAA28673@www.ccl.net> Date: Mon, 31 Oct 94 08:24 CST Subject: Convergence in MOPAC C.I. calculations To: CHEMISTRY@ccl.net X-VMS-To: IN%"CHEMISTRY@ccl.net" Hello, I am using MOPAC 6.0 to do calculations on fairly large polycyclic molecules related to porphyrins. Part of this work involves optimizing geometries and calculating energies for the cations and anions of these molecules. To make these calculations more accurate (hopefully) we are using the configuration interaction options in MOPAC 6, but have encountered some problems with convergence. After reducing the gradient to below 1.0, the optimization begins to oscillate, and shows no sign of convergence after running for several hours. A typical command line is as follows: AM1 PRECISE CHARGE=-1 C.I.=(4,2) VECTORS NODIIS We have also tried the eigenvector following converger with similar results. If anyone has had experience with a similar problem, I would appreciate any advice that you could provide on this subject. The manual is a bit brief on both configuration interaction calculations and convergence problems, so I thought I might try here. You can mail me directly and I will summarize the responses to the list if there is interest. Thanks in advance.... Kristopher Wise University of Oklahoma Department of Chemistry and Biochemistry KEWISE@aardvark.ucs.uoknor.edu From noy@tci005.uibk.ac.at Mon Oct 31 10:34:53 1994 Received: from uibk.ac.at for noy@tci005.uibk.ac.at by www.ccl.net (8.6.9/930601.1506) id JAA28901; Mon, 31 Oct 1994 09:38:27 -0500 Received: from tci005.uibk.ac.at by uibk.ac.at with SMTP id AA21287 (5.65c/IDA-1.4.4 for ); Mon, 31 Oct 1994 15:37:52 +0100 Received: by tci005.uibk.ac.at (AIX 3.2/UCB 5.64/CFIBK-2e.AIX) id AA23668; Mon, 31 Oct 1994 15:37:54 +0100 From: noy@tci005.uibk.ac.at (Noy) Message-Id: <9410311437.AA23668@tci005.uibk.ac.at> Subject: 100% parallel MD (1st Summary) To: chemistry@ccl.net Date: Mon, 31 Oct 1994 15:37:53 +0100 (NFT) Cc: supot@atc.atccu.chula.ac.th (Dr. Supot Hannongbua), nan@atc.atccu.chula.ac.th (Anan Tongra-ar), ypp@tci002.uibk.ac.at X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 3567 Dear fellows, Thanks a lot for your kind suggestions and comments. I have summarized the response. All of the responses will be posted on the next mail. Many responses state that the connection of two MD trajectories are in doubt, due to numerical error in integrating schemes. Some of them believe that the reversibility of velocity is disturbed by the coupling with Thermo-bath and the cut-off of the potential. However, a few people agree on this idea. My point of view is that the structural properties obtained from connecting two MD trajectories are the same as running one MD double time. t0------------->t1 t2 <---------------t0 and t0--------------------------------->t1 If we look at the phase space (p,q) p | | ta \ | \ | \ | \ ....... t1 | \ t0 ......./ | ....+ | / | ...../ | *' | t2 | |___________________________________________ q Well, I start equilibrating the system from the time ta up to the time t0, which I am sure it is an equilibrium state. Then at this point, I traditionally perform a simulation from time t0 to the time t1. At the time t0, I revert the sign of all velocities and simulate in parallel up to the time t2. I believe that both of the trajectories from the time t0--->t1 and t0--->t2 are in equilibrium state. So, the phase space should be expored correctly, isn't it ? The statistical average shoud be, hence, conserved. If I revert all momentum back again from the time t2--->t0, then two trajectories are connected to produce one longer run. My purpose is to have longer time-correlation rather than having more statistical samples. So, I must be sure that this is correct or not. Many people are in doubt that the reconnection of two MD is invalid. I would like to ask some comments if it is possible or not that ...... Since the prediction of the next coordinates and velocities are respective to the current (t) and the former (t-1) one. If I have this in hand t, t-1 and t+1 t = t0 I integrate forwardly from t0 ----> t1 using t-1 as the previous step and I integrate backwardly from t0----> t2 using t+1 as the previous step. Do you think the connectiong is smooth and neat ? As stated by many people that the past of the trajectory is lost due to the numerical error in integration scheme, I would believe so. Anyway, if both trajectories are still in the equilibrium state, they should represent the phase space correctly. Some ideas ? my best regards, Teerakiat ---------------------------------------------------------------------------- Teerakiat Kerdcharoen (NOY) Institute of General and Inorganic and Theoretical Chemistry Innrain 52a, A-6020 Innsbruck AUSTRIA e-mail: noy@tci2.uibk.ac.at, noy@tci.uibk.ac.at, c72454@cx.uibk.ac.at : noy@atc.atccu.chula.ac.th, noy@atc2.atccu.chula.ac.th ( Bangkok ) Research : Molecular Dynamics simulations : Computer Aided Molecular/Material Designs ----------------------------------------------------------------------------- *** I have no past and no future. I just have today. From rsefeck@watson.ibm.com Mon Oct 31 12:22:32 1994 Received: from watson.ibm.com for rsefeck@watson.ibm.com by www.ccl.net (8.6.9/930601.1506) id MAA02756; Mon, 31 Oct 1994 12:14:37 -0500 From: Message-Id: <199410311714.MAA02756@www.ccl.net> Received: from YKTVMV by watson.ibm.com (IBM VM SMTP V2R3) with BSMTP id 1793; Mon, 31 Oct 94 12:14:42 EST Date: Mon, 31 Oct 94 11:43:36 EST To: chemistry@ccl.net Subject: J.P.Stewart number Does anyone know how to get in touch with Dr. J. P. Stewart?(MOPAC) I understand he is no longer at the Airforce Labs and that he may be at Fujitsu. But with 4 separate phone calls to Fujitsu no one has heard of him. Any help would be appreciated. Rich Sefecka IBM Watson Labs email me direct at RSEFECK@watson.ibm.com thanks From kottalam@rhea.cray.com Mon Oct 31 12:27:12 1994 Received: from timbuk.cray.com for kottalam@rhea.cray.com by www.ccl.net (8.6.9/930601.1506) id LAA02349; Mon, 31 Oct 1994 11:52:57 -0500 Received: from wake.cray.com (kottalam@wake.cray.com [128.162.172.2]) by timbuk.cray.com (8.6.9/8.6.9N) with SMTP id KAA26079 for ; Mon, 31 Oct 1994 10:52:56 -0600 Received: by wake.cray.com (4.1/CRI-5.14) id AA04798; Mon, 31 Oct 94 10:52:54 CST Date: Mon, 31 Oct 94 10:52:54 CST From: kottalam@rhea.cray.com (Jeyapandian Kottalam) Message-Id: <9410311652.AA04798@wake.cray.com> To: chemistry@ccl.net Subject: Re: CCL:100% parallel MD > From: Martyn Winn > > I wonder if anyone is answering the original question > about efficiency?! I think the efficiency question did not evoke much discussion because everyone would agree with the theoretical 100% parallelism. However, the time-parallelism is not very interesting because it is strictly limited to a factor of 2. From the Massively Parallel point of view, we would like to have scalability to 500 to 1000 processors even with slightly less than 100% parallelism. This factor of 2 limitation is in addition to (or in spite of) the chaotic nature of MD trajectories. Kottalam Phone: (USA) 612 683 3622 Cray Research, Inc. From sunger@crl.com Mon Oct 31 12:30:17 1994 Received: from mail.crl.com for sunger@crl.com by www.ccl.net (8.6.9/930601.1506) id LAA02058; Mon, 31 Oct 1994 11:40:15 -0500 Received: from crl5.crl.com by mail.crl.com with SMTP id AA08433 (5.65c/IDA-1.5 for ); Mon, 31 Oct 1994 08:24:10 -0800 Received: by crl5.crl.com id AA10026 (5.65c/IDA-1.5); Mon, 31 Oct 1994 08:23:22 -0800 Date: Mon, 31 Oct 1994 08:23:22 -0800 (PST) From: Stefan Unger To: Dimitris Agrafiotis Cc: chemistry@ccl.net Subject: Re: CCL:Need C/C++ libraries for clustering In-Reply-To: <9410281714.ZM10482@europa.3dp.com> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII You might try contacting Daylight Chemical (Irvine, CA). They have worked on just this problem. Call Yossi Taitz 714-476-0451. Good luck. Stefan On Fri, 28 Oct 1994, Dimitris Agrafiotis wrote: > I am looking for public/commercial C/C++ libraries for clustering large > amounts of multidimensional data. Your help would be greatly appreciated. > I will summarize and submit the responses to the list. Thanks for your > help. > > > > -- > Dimitris K. Agrafiotis, PhD | e-mail: dimitris@3dp.com > 3-Dimensional Pharmaceuticals, Inc. | tel: (215) 222-8950 > 3700 Market Street | fax: (215) 222-8960 > Philadelphia, PA 19104 > > > > ---Administrivia: This message is automatically appended by the mail exploder: > CHEMISTRY@ccl.net -- everyone | CHEMISTRY-REQUEST@ccl.net -- coordinator > MAILSERV@ccl.net: HELP CHEMISTRY | Gopher: www.ccl.net 73 > Anon. ftp www.ccl.net | CHEMISTRY-SEARCH@ccl.net -- archive search > http://www.ccl.net/chemistry.html | for info send: HELP SEARCH to MAILSERV > > From jshaffer@anemone.usc.edu Mon Oct 31 13:22:35 1994 Received: from usc.edu for jshaffer@anemone.usc.edu by www.ccl.net (8.6.9/930601.1506) id NAA03779; Mon, 31 Oct 1994 13:03:56 -0500 Received: from anemone.usc.edu by usc.edu (4.1/SMI-3.0DEV3-USC+3.1) id AA15635; Mon, 31 Oct 94 10:03:53 PST Received: (jshaffer@localhost) by anemone.usc.edu (8.6.8.1/8.6.7+ucs) id KAA22945; Mon, 31 Oct 1994 10:03:54 -0800 Date: Mon, 31 Oct 1994 10:03:54 -0800 From: J Scott Shaffer Message-Id: <199410311803.KAA22945@anemone.usc.edu> To: chemistry@ccl.net Subject: Re: 100% parallel MD Reply-To: jshaffer@che1.usc.edu In my previous post, I botched the graph. The phase point would be reflected through Q=0 upon reversing the velocities, giving something like: P | | | ta \ | \ | \ | \ t1 | \ t0 ......./ | ....+ | / | ...../ | *' | t2 | ______________|___________________________________ Q | | | | | | | t0' | . | . | t2' . | . . | ..... | | | Sorry for the wasted space. Scott From jshaffer@anemone.usc.edu Mon Oct 31 13:24:08 1994 Received: from usc.edu for jshaffer@anemone.usc.edu by www.ccl.net (8.6.9/930601.1506) id MAA03617; Mon, 31 Oct 1994 12:50:55 -0500 Received: from anemone.usc.edu by usc.edu (4.1/SMI-3.0DEV3-USC+3.1) id AA15102; Mon, 31 Oct 94 09:50:52 PST Received: (jshaffer@localhost) by anemone.usc.edu (8.6.8.1/8.6.7+ucs) id JAA22939; Mon, 31 Oct 1994 09:50:52 -0800 Date: Mon, 31 Oct 1994 09:50:52 -0800 From: J Scott Shaffer Message-Id: <199410311750.JAA22939@anemone.usc.edu> To: chemistry@ccl.net Subject: Re: 100% parallel MD (1st summary) Reply-To: jshaffer@che1.usc.edu Dear Dr. Kerdcharoen: I have entered this discussion a little late, but here goes. Song Ling has presented these arguments earlier; here they are again in a slightly different form. Let's talk only about integrating Newton's equations, yielding MD simulations in the NVE ensemble. There are two points to be made: (1) Disregarding numerical error in the time integration, to generate continuous trajectories, one must reverse the sign of the time step, dt ----> -dt, keeping the same values of the positions *and* velocities. Correct time correlation functions cannot be obtained by reversing the signs of velocities and connecting the two resulting trajectories. By reversing the velocities, you change the location of the system in phase space, as illustrated below, by reflecting the momenta through the P=0 axis. Thus, any time correlation functions that are calculated by joining the two trajectories would be incorrect, since the trajectories are not continuous in phase space. The phase space (Q,P): The points ta, t0, t1, and t2 are those plotted by Dr. Kerdcharoen in his first summary. The point t0' is the phase point obtained by reversing the velocities of the system at point t0; the point t2' is the final location of the system in phase space after integrating Newton's equations from point t0'. P | | | ta \ | \ | \ | \ t1 t0' | \ t0 ......./ t2' . | ....+ . . . | / . . . | ...../ . . . . | *' | t2 | ______________________________|___________________________________ Q (2) With numerical error in the integration of the equations of motion, no trajectory is reversible in practice. This is illustrated nicely in Figure 3.1 of Allen and Tildesley, "Computer Simulation of Liquids", where they show that trajectories that should in principle be identical actually diverge from each other rapidly because of numerical error in the time integration. This is not necessarily a problem in terms of calculating ensemble averages from MD. In fact, if one were to use the dt ----> -dt reversal described in (1), this numerical error would be *essential*, otherwise you would end up back in the unequilibrated starting configuration! Scott ===================================================================== J. Scott Shaffer jshaffer@che1.usc.edu Department of Chemical Engineering University of Southern California (213) 740-2067 Los Angeles, CA 90089-1211 USA (213) 740-8053 (FAX) ===================================================================== From topper@cooper.edu Mon Oct 31 13:25:17 1994 Received: from magnum.cooper.edu for topper@cooper.edu by www.ccl.net (8.6.9/930601.1506) id MAA03608; Mon, 31 Oct 1994 12:49:01 -0500 Received: from zeus (zeus.cooper.edu) by magnum.cooper.edu with SMTP id AA20271 (5.65c/IDA-1.4.4 for ); Mon, 31 Oct 1994 12:49:20 -0500 Received: by zeus id AA15315 (5.67b/IDA-1.5); Mon, 31 Oct 1994 12:47:57 -0500 Date: Mon, 31 Oct 1994 12:47:57 -0500 From: Robert_Topper Message-Id: <199410311747.AA15315@zeus> To: CHEMISTRY@ccl.net, noy@tci005.uibk.ac.at Subject: Re: CCL:100% parallel MD Content-Length: 6112 BANDWIDTH ALERT: A LONG discussion on chaos in molecular systems follows. I hope that some find it interesting. - robert > > Two points: > 1) My impression is that there is much work devoted to > chaotic trajectories for small molecules (i.e. with a > small no. of degrees of freedom) but that the issue is > ignored for MD simulations of bulk systems. This is, of > course, purely for reasons of convenience. Yes and no...certainly most work in nonlinear dynamics and chaos of MD trajectories has concentrated on small molecular models at fixed E, but not simply for "reasons of convenience." The fact of the matter is that there remains a lot to learn about nonlinear dynamics in conservative Hamiltonian systems, esp. in high-dimensional spaces. There are nonlinear phenonema that only occur in systems with dimension >/= 3 (ex: Arnold diffusion)...plus the beautiful visualization tools suggested by Poincare (Poincare maps, or "surfaces of section") are not unique phase-space maps in >/= 3 dimensions. Progress into understanding chaos in "big" systems has been slow (but see recent papers in Physica D and JCP for counterexamples). I personally believe that the problems are less numerical than fundamental, and more basic theory needs to emerge to make some progress. Song Ling has just authored a series of interesting papers on HCN isomerization and chaos (although he's too modest to tell you so). I did my thesis research in this area. I'd be glad to steer those who are interested towards a few key papers. > 2) If one is in a chaotic regime, then I'm sure that > the -dt, dt trajectory will differ from the 2dt trajectory, > but taken individually each would be a perfectly valid > trajectory. There would be no reason to prefer one over > the other. In fact, a chaotic trajectory is surely ideal > to satisfy the ergodic requirements of most MD simulations. > > I wonder if anyone is answering the original question > about efficiency?! Martyn has an excellent point. The "practical" consideration for those interested in getting at the NVT or NpT ensembles (either directly or otherwise) is: am I sampling "properly" and "completely" (i.e., does my sample accurately represent the statistical properties of the distribution function I'm trying to mimic, and do I have enough configurations)? This doesn't necessarily mean that the trajectory needs to be DYNAMICALLY correct...if you want only THERMODYNAMIC properties. However, if you also want dynamic properties (autocorrelation functions?), the distinction between a (-dt,dt) trajectory and a (0,2t) trajectory could be relevant. I think this should be ironed out before sweating out how to make the calculation efficient. That said, I personally don't believe that whether the trajectory is chaotic or not is relevant here. The Lyapunov exponent for the starting point in phase space will give the same "divergence rate" whether one integrates asymmetrically (0,2t) or symmetrically (-t, t)... either numerically generated trajectory is just as "wrong" (or just as "right" depending on your point of view). Chaos theory tells us that it's fundamentally impossible to exactly GENERATE a chaotic trajectory...unless you computer carries an infinite number of significant figures. Fundamentally, your ability to back-integrate on the interval (2t,0) is going to be just as poor (or good) as your ability to back-integrate (t,-t). If the trajectory is chaotic, it's chaotic. Chaos / nonlinear dynamics has some important things to say about the calculation of thermodynamic properties using MD in terms of satisfaction / nonsatisfaction of the Ergodic Hypothesis. Constant-(NVE) computer experiments on many-oscillator systems carried out by Fermi/Pasta/Ulam at the dawn of "computational experimentation" showed that quasiperiodic volumes of phase space can be sizeable no matter how many particles you have in the system. There are volumes of phase space which may have non-zero "measure" and which are (at least on the time scales of simulations) toroidal in shape. Chaotic trajectories are what one hopes for in MD...they come closest to satisfying the ergodic hypothesis, because they sample a lot of phase space (in principle, they will sample all of phase space given infinite time). However, even a chaotic trajectory cannot sample the ENTIRE phase space on an MD time scale IF there are "regular" regions of phase space...because such regions are not accessible to a chaotic trajectory, they are excluded from it (until Arnold diffusion, a very very slow process, sets in). So one needs ways to "skip" between regular and chaotic regions of phase space in one's sampling...sometimes. This is really probably only an issue IF the regular regions have big volumes, and EITHER (1) These "excluded phase-space volumes" lock out significant portions of CONFIGURATION space, and/or (2) one is averaging something which depends on configurations AND/OR MOMENTA (i.e., your averaged quantity is defined in the full phase space). Usually in thermodynamics one looks at properties which are solely configuration-space averages, so (2) is not often the case. Of course the "masters of MD" (you know who you are) have devised various sampling tricks over the years to ensure that (1) and (2) are not a problem. But when calculating phase-space averages in the microcanonical ensemble, one must be aware of the potential for trouble if there are sizeable and/or numerous "islands" of quasiperiodicity in the "chaotic sea." I hope that this is helpful or interesting - awaiting inevitable flames robert ************************************************************************ Prof. Robert Q. Topper internet: topper@cooper.edu Department of Chemistry phone: (212) 353-4378 The Cooper Union FAX: (212) 353-4341 Cooper Square subway: take the 6 to Astor Place New York, NY 10003 USA or the N/R to 8th St/NYU ************************************************************************ Thermodynamics forever... From bennett@ubaclu.unibas.ch Mon Oct 31 13:26:24 1994 Received: from swiba9.unibas.ch for bennett@ubaclu.unibas.ch by www.ccl.net (8.6.9/930601.1506) id NAA03883; Mon, 31 Oct 1994 13:13:09 -0500 Received: from yogi.urz.unibas.ch by swiba9.unibas.ch with SMTP (PP); Mon, 31 Oct 1994 19:12:59 +0100 Received: from fib.chemie.unibas.ch. (fib.chemie.unibas.ch) by ubaclu.unibas.ch (PMDF V4.3-7 #7496) id <01HIXQYCAA4099DJUW@ubaclu.unibas.ch>; Mon, 31 Oct 1994 19:12:54 +0100 Date: Mon, 31 Oct 1994 19:11:53 +0100 From: Frederick Bennett Subject: C60 Z matrices summary To: chemistry@ccl.net Message-id: X-Mailer: VersaTerm Link v1.1 Content-transfer-encoding: 7BIT Dear to whom it may concern, some time ago I posted a request for Z matrices for C60 that observed the molecular point group of the species. I am posting a summary for those who are interested and for the sake of providing a compilation for the CCL archives the replies a listed below. Regards Frederick Bennett =============================================================================== David Heisterberg (djh@ccl.net) X70 X71 X70 3.50 X72 X70 3.50 X71 180.00 X73 X70 3.50 X71 63.4349488229221 X72 0.00 X77 X70 3.50 X71 63.4349488229221 X73 72.00 X76 X70 3.50 X71 63.4349488229221 X73 144.00 X75 X70 3.50 X71 63.4349488229221 X73 -144.00 X74 X70 3.50 X71 63.4349488229221 X73 -72.00 X78 X70 3.50 X71 116.565051177078 X73 36.00 X79 X70 3.50 X71 116.565051177078 X73 108.00 X80 X70 3.50 X71 116.565051177078 X73 180.00 X82 X70 3.50 X71 116.565051177078 X73 -108.00 X81 X70 3.50 X71 116.565051177078 X73 -36.00 C2 X71 1.236 X70 80.00 X74 0.00 C9 X71 1.236 X70 80.00 X74 72.00 C18 X71 1.236 X70 80.00 X74 144.00 C13 X71 1.236 X70 80.00 X74 -72.00 C55 X71 1.236 X70 80.00 X74 -144.00 C10 X72 1.236 X70 80.00 X79 0.00 C34 X72 1.236 X70 80.00 X79 72.00 C44 X72 1.236 X70 80.00 X79 144.00 C6 X72 1.236 X70 80.00 X79 -72.00 C29 X72 1.236 X70 80.00 X79 -144.00 C3 X75 1.236 X70 80.00 X71 0.00 C7 X75 1.236 X70 80.00 X71 72.00 C22 X75 1.236 X70 80.00 X71 144.00 C17 X75 1.236 X70 80.00 X71 -72.00 C37 X75 1.236 X70 80.00 X71 -144.00 C20 X76 1.236 X70 80.00 X71 0.00 C19 X76 1.236 X70 80.00 X71 72.00 C38 X76 1.236 X70 80.00 X71 144.00 C57 X76 1.236 X70 80.00 X71 -72.00 C59 X76 1.236 X70 80.00 X71 -144.00 C56 X77 1.236 X70 80.00 X71 0.00 C58 X77 1.236 X70 80.00 X71 72.00 C28 X77 1.236 X70 80.00 X71 144.00 C54 X77 1.236 X70 80.00 X71 -72.00 C27 X77 1.236 X70 80.00 X71 -144.00 C15 X73 1.236 X70 80.00 X71 0.00 C53 X73 1.236 X70 80.00 X71 72.00 C51 X73 1.236 X70 80.00 X71 144.00 C16 X73 1.236 X70 80.00 X71 -72.00 C49 X73 1.236 X70 80.00 X71 -144.00 C8 X74 1.236 X70 80.00 X71 0.00 C14 X74 1.236 X70 80.00 X71 72.00 C47 X74 1.236 X70 80.00 X71 144.00 C1 X74 1.236 X70 80.00 X71 -72.00 C21 X74 1.236 X70 80.00 X71 -144.00 C12 X78 1.236 X70 80.00 X72 0.00 C30 X78 1.236 X70 80.00 X72 72.00 C52 X78 1.236 X70 80.00 X72 144.00 C5 X78 1.236 X70 80.00 X72 -72.00 C25 X78 1.236 X70 80.00 X72 -144.00 C4 X79 1.236 X70 80.00 X72 0.00 C11 X79 1.236 X70 80.00 X72 72.00 C26 X79 1.236 X70 80.00 X72 144.00 C33 X79 1.236 X70 80.00 X72 -72.00 C60 X79 1.236 X70 80.00 X72 -144.00 C36 X80 1.236 X70 80.00 X72 0.00 C35 X80 1.236 X70 80.00 X72 72.00 C40 X80 1.236 X70 80.00 X72 144.00 C42 X80 1.236 X70 80.00 X72 -72.00 C39 X80 1.236 X70 80.00 X72 -144.00 C43 X82 1.236 X70 80.00 X72 0.00 C41 X82 1.236 X70 80.00 X72 72.00 C24 X82 1.236 X70 80.00 X72 144.00 C45 X82 1.236 X70 80.00 X72 -72.00 C23 X82 1.236 X70 80.00 X72 -144.00 C31 X81 1.236 X70 80.00 X72 0.00 C46 X81 1.236 X70 80.00 X72 72.00 C48 X81 1.236 X70 80.00 X72 144.00 C32 X81 1.236 X70 80.00 X72 -72.00 C50 X81 1.236 X70 80.00 X72 -144.00 ======================================================================== E. Lewars (elewars@alchemy.chem.utoronto.ca) 1994 Sept 18 Frederick Bennett asked for a Z-matrix or some other info for the geometry of C60. Here are the Cartesians; they seem to have the exact correct symmetry. It should be possible to use them to generate a Z-matrix, e.g. using MOPAC's "0scf aigout noxyz" keywords. although I haven't tried this with C60. 60 Buckyball C 3.451266498 0.685000000 0.000000000 C 3.451266498 -0.685000000 0.000000000 C -3.451266498 0.685000000 0.000000000 C -3.451266498 -0.685000000 0.000000000 C 0.685000000 0.000000000 3.451266498 C -0.685000000 0.000000000 3.451266498 C 0.685000000 0.000000000 -3.451266498 C -0.685000000 0.000000000 -3.451266498 C 0.000000000 3.451266498 0.685000000 C 0.000000000 3.451266498 -0.685000000 C 0.000000000 -3.451266498 0.685000000 C 0.000000000 -3.451266498 -0.685000000 C 3.003809890 1.409000000 1.171456608 C 3.003809890 1.409000000 -1.171456608 C 3.003809890 -1.409000000 1.171456608 C 3.003809890 -1.409000000 -1.171456608 C -3.003809890 1.409000000 1.171456608 C -3.003809890 1.409000000 -1.171456608 C -3.003809890 -1.409000000 1.171456608 C -3.003809890 -1.409000000 -1.171456608 C 1.409000000 1.171456608 3.003809890 C 1.409000000 -1.171456608 3.003809890 C -1.409000000 1.171456608 3.003809890 C -1.409000000 -1.171456608 3.003809890 C 1.409000000 1.171456608 -3.003809890 C 1.409000000 -1.171456608 -3.003809890 C -1.409000000 1.171456608 -3.003809890 C -1.409000000 -1.171456608 -3.003809890 C 1.171456608 3.003809890 1.409000000 C -1.171456608 3.003809890 1.409000000 C 1.171456608 3.003809890 -1.409000000 C -1.171456608 3.003809890 -1.409000000 C 1.171456608 -3.003809890 1.409000000 C -1.171456608 -3.003809890 1.409000000 C 1.171456608 -3.003809890 -1.409000000 C -1.171456608 -3.003809890 -1.409000000 C 2.580456608 0.724000000 2.279809890 C 2.580456608 0.724000000 -2.279809890 C 2.580456608 -0.724000000 2.279809890 C 2.580456608 -0.724000000 -2.279809890 C -2.580456608 0.724000000 2.279809890 C -2.580456608 0.724000000 -2.279809890 C -2.580456608 -0.724000000 2.279809890 C -2.580456608 -0.724000000 -2.279809890 C 0.724000000 2.279809890 2.580456608 C 0.724000000 -2.279809890 2.580456608 C -0.724000000 2.279809890 2.580456608 C -0.724000000 -2.279809890 2.580456608 C 0.724000000 2.279809890 -2.580456608 C 0.724000000 -2.279809890 -2.580456608 C -0.724000000 2.279809890 -2.580456608 C -0.724000000 -2.279809890 -2.580456608 C 2.279809890 2.580456608 0.724000000 C -2.279809890 2.580456608 0.724000000 C 2.279809890 2.580456608 -0.724000000 C -2.279809890 2.580456608 -0.724000000 C 2.279809890 -2.580456608 0.724000000 C -2.279809890 -2.580456608 0.724000000 C 2.279809890 -2.580456608 -0.724000000 C -2.279809890 -2.580456608 -0.724000000 =========================================================================== Jorge (JSMCM@jazz.ucc.uno.edu Here is a Z-matrix for C60 that I made some time ago for G92. The problem that my copy of G92 can not handle the Ih symmetry and default to C1 (no symmetry). The exact value of the angle a is 63.43494886 By truncating the value of the angle a to 63.43 you can force G92 to use D5D symmetry Note: you must fix the two angles a and w. you can optimized distances f and d only f is the distance from the center of the C60 to the center of any pentagon d is the distance fron the center of pentagon to any of its Carbons. x1 x2 1 f x3 1 f 2 a x4 1 f 2 a 3 w x5 1 f 2 a 4 w x6 1 f 2 a 5 w x7 1 f 2 a 6 w x8 1 1.0 2 90.0 3 0.0 x9 1 f 8 90.0 2 180.0 x10 1 f 9 a 8 180.0 x11 1 f 9 a 10 w x12 1 f 9 a 11 w x13 1 f 9 a 12 w x14 1 f 9 a 13 w c1a x2 d x1 90.0 x8 0.0 c2a x2 d x1 90.0 c1a 72.0 c3a x2 d x1 90.0 c2a 72.0 c4a x2 d x1 90.0 c3a 72.0 c5a x2 d x1 90.0 c4a 72.0 c1b x3 d x1 90.0 x2 0.0 c2b x3 d x1 90.0 c1b 72.0 c3b x3 d x1 90.0 c2b 72.0 c4b x3 d x1 90.0 c3b 72.0 c5b x3 d x1 90.0 c4b 72.0 c1c x4 d x1 90.0 x2 0.0 c2c x4 d x1 90.0 c1c 72.0 c3c x4 d x1 90.0 c2c 72.0 c4c x4 d x1 90.0 c3c 72.0 c5c x4 d x1 90.0 c4c 72.0 c1d x5 d x1 90.0 x2 0.0 c2d x5 d x1 90.0 c1d 72.0 c3d x5 d x1 90.0 c2d 72.0 c4d x5 d x1 90.0 c3d 72.0 c5d x5 d x1 90.0 c4d 72.0 c1e x6 d x1 90.0 x2 0.0 c2e x6 d x1 90.0 c1e 72.0 c3e x6 d x1 90.0 c2e 72.0 c4e x6 d x1 90.0 c3e 72.0 c5e x6 d x1 90.0 c4e 72.0 c1f x7 d x1 90.0 x2 0.0 c2f x7 d x1 90.0 c1f 72.0 c3f x7 d x1 90.0 c2f 72.0 c4f x7 d x1 90.0 c3f 72.0 c5f x7 d x1 90.0 c4f 72.0 c1g x9 d x1 90.0 x8 180.0 c2g x9 d x1 90.0 c1g 72.0 c3g x9 d x1 90.0 c2g 72.0 c4g x9 d x1 90.0 c3g 72.0 c5g x9 d x1 90.0 c4g 72.0 c1h x10 d x1 90.0 x9 0.0 c2h x10 d x1 90.0 c1h 72.0 c3h x10 d x1 90.0 c2h 72.0 c4h x10 d x1 90.0 c3h 72.0 c5h x10 d x1 90.0 c4h 72.0 c1i x11 d x1 90.0 x9 0.0 c2i x11 d x1 90.0 c1i 72.0 c3i x11 d x1 90.0 c2i 72.0 c4i x11 d x1 90.0 c3i 72.0 c5i x11 d x1 90.0 c4i 72.0 c1j x12 d x1 90.0 x9 0.0 c2j x12 d x1 90.0 c1j 72.0 c3j x12 d x1 90.0 c2j 72.0 c4j x12 d x1 90.0 c3j 72.0 c5j x12 d x1 90.0 c4j 72.0 c1k x13 d x1 90.0 x9 0.0 c2k x13 d x1 90.0 c1k 72.0 c3k x13 d x1 90.0 c2k 72.0 c4k x13 d x1 90.0 c3k 72.0 c5k x13 d x1 90.0 c4k 72.0 c1l x14 d x1 90.0 x9 0.0 c2l x14 d x1 90.0 c1l 72.0 c3l x14 d x1 90.0 c2l 72.0 c4l x14 d x1 90.0 c3l 72.0 c5l x14 d x1 90.0 c4l 72.0 f=3.32182409 d=1.24427177 a=63.43[494886] [truncation will force D5D symm] w=72.0 enjoy Jorge ============================================================================ Dr. James Stewart (jstewart@fujitsu.com Here is a MOPAC Z-matrix for C60. There are two unknowns, as required by symmetry. James J. P. Stewart XX 0.0000000 0 0.000000 0 0.000000 0 0 0 0 XX 3.4932251 1 0.000000 0 0.000000 0 1 0 0 XX 3.4932243 0 108.000000 0 0.000000 0 1 2 0 XX 3.4932243 0 108.000000 0 0.000000 0 2 1 3 XX 3.4932243 0 108.000000 0 63.434949 0 2 1 3 XX 3.4932243 0 108.000000 0 -63.434949 0 1 2 3 XX 3.4932243 0 108.000000 0 -63.434949 0 1 6 2 XX 3.4932243 0 108.000000 0 63.434949 0 1 3 2 XX 3.4932243 0 108.000000 0 0.000000 0 5 2 1 XX 3.4932243 0 108.000000 0 0.000000 0 3 1 2 XX 3.4932243 0 108.000000 0 -63.434949 0 3 1 2 XX 3.4932243 0 108.000000 0 63.434949 0 4 2 1 C 1.2398064 1 90.000000 0 0.000000 0 1 12 6 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 1 12 13 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 1 12 13 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 1 12 13 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 1 12 13 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 2 11 5 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 2 11 18 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 2 11 18 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 2 11 18 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 2 11 18 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 3 5 6 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 3 5 23 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 3 5 23 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 3 5 23 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 3 5 23 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 4 6 5 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 4 6 28 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 4 6 28 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 4 6 28 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 4 6 28 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 5 3 4 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 5 3 33 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 5 3 33 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 5 3 33 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 5 3 33 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 6 4 1 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 6 4 38 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 6 4 38 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 6 4 38 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 6 4 38 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 7 9 1 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 7 9 43 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 7 9 43 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 7 9 43 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 7 9 43 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 8 10 1 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 8 10 48 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 8 10 48 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 8 10 48 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 8 10 48 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 9 7 5 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 9 7 53 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 9 7 53 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 9 7 53 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 9 7 53 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 10 8 11 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 10 8 58 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 10 8 58 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 10 8 58 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 10 8 58 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 11 2 9 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 11 2 63 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 11 2 63 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 11 2 63 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 11 2 63 0.0000 C 1.2398104 0 90.000000 0 0.000000 0 12 1 4 0.0000 C 1.2398104 0 90.000000 0 72.000000 0 12 1 68 0.0000 C 1.2398104 0 90.000000 0 144.000000 0 12 1 68 0.0000 C 1.2398104 0 90.000000 0 -144.000000 0 12 1 68 0.0000 C 1.2398104 0 90.000000 0 -72.000000 0 12 1 68 0.0000 ====================================================== Papernet Address : Frederick R. Bennett Institute for Physical Chemistry Klingelbergstrasse 80 CH-4056 Basel Switzerland Mouthnet Address : [41]-(61)-267-3821 Internet Address : bennett@ubaclu.unibas.ch ====================================================== From rsefeck@watson.ibm.com Mon Oct 31 14:22:38 1994 Received: from watson.ibm.com for rsefeck@watson.ibm.com by www.ccl.net (8.6.9/930601.1506) id OAA05034; Mon, 31 Oct 1994 14:05:12 -0500 From: Message-Id: <199410311905.OAA05034@www.ccl.net> Received: from YKTVMV by watson.ibm.com (IBM VM SMTP V2R3) with BSMTP id 6383; Mon, 31 Oct 94 14:05:18 EST Date: Mon, 31 Oct 94 14:02:11 EST To: chemistry@ccl.net Subject: J.P. Stewart Thanks to all the members of this forum for the 32 responses I received on how to contact Dr. Stewart. Rich From szeinfel@snfma1.if.usp.br Mon Oct 31 15:22:41 1994 Received: from snfma1.if.usp.br for szeinfel@snfma1.if.usp.br by www.ccl.net (8.6.9/930601.1506) id PAA06363; Mon, 31 Oct 1994 15:10:36 -0500 Received: (szeinfel@localhost) by snfma1.if.usp.br (8.6.8/8.6.4) id SAA05773; Mon, 31 Oct 1994 18:09:47 -0200 Date: Mon, 31 Oct 1994 18:09:46 -0200 (EDT) From: Rafael Iosef Najmanovich Szeinfeld {S To: computational chemistry Subject: postscript papers. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi Netter's, Does someone knows if there are any available internet addresses where we can find preprint papers or already published papers in postscript form, something like the high-energy-physics preprint service ? Thank's in advance, Rafael. *-----------------------------------------------------------------------------* * Rafael Iosef Najmanovich Szeinfeld |SMAIL: Depto. de Bioquimica - B10 INF.* * Dept. Biochemistry -Chemistry Inst. | Universidade de Sao Paulo * * Dept. Math. Physics -Physics Inst. | Av. Prof. Lineu Prestes 748 * * University of Sao Paulo | CEP 05508-900 * * E-MAIL : szeinfel@snfma1.if.usp.br | Sao Paulo - SP - Brazil * *-----------------------------------------------------------------------------* From d3g359@rahman.pnl.gov Mon Oct 31 17:22:36 1994 Received: from pnl.gov for d3g359@rahman.pnl.gov by www.ccl.net (8.6.9/930601.1506) id RAA08380; Mon, 31 Oct 1994 17:06:43 -0500 From: Received: from carbon.pnl.gov by pnl.gov (PMDF V4.2-15 #4032) id <01HIXG8SIK5S00DICL@pnl.gov>; Mon, 31 Oct 1994 14:05:57 PST Received: from pnlg.pnl.gov by carbon.pnl.gov (4.1/SMI-4.1) id AA00477; Mon, 31 Oct 94 14:05:45 PST Received: from rahman.pnl.gov (130.20.182.20) by pnlg.pnl.gov; Mon, 31 Oct 94 13:56 pst Received: by rahman.pnl.gov (931110.SGI/920502.SGI) for @carbon.pnl.gov:CHEMISTRY@ccl.net id AA17639; Mon, 31 Oct 94 13:56:23 -0800 Date: Mon, 31 Oct 1994 13:56:23 -0800 Subject: Simulated IR spectra To: CHEMISTRY@ccl.net Message-id: <9410312156.AA17639@rahman.pnl.gov> Content-transfer-encoding: 7BIT All, I have seen a few papers in which an IR spectra is simulated by a sum of gaussian (or triangular) functions based on IR frequencies and intensities from a QM program such as G92. Obviously, this is a crude approximation, since the width of the gaussians can only be guessed. Still, it seems to be "useful" to compare to experiment. Does anyone have opinions about this? Does anyone have a "rule of thumb" about the width of the gaussians? Does anyone have a program that already does this? Thanks in advance. John ------------------------------------------------------------------------------ John Nicholas Office: (509) 375-6559 Senior Research Scientist FAX: (509) 375-6631 Molecular Science Research Center Pacific Northwest Laboratory Mailstop K1-90 Richland, WA 99352 ------------------------------------------------------------------------------ From SCHUBERT%cmda@randb.abbott.com Mon Oct 31 17:30:03 1994 Received: from abtlabs.abbott.com for SCHUBERT%cmda@randb.abbott.com by www.ccl.net (8.6.9/930601.1506) id QAA07820; Mon, 31 Oct 1994 16:40:32 -0500 From: Received: from randb.pprd.abbott.com (randb.abbott.com) by abtlabs.abbott.com with SMTP id AA07588 (5.65c/IDA-1.4.4 for ); Mon, 31 Oct 1994 15:37:32 -0600 Received: from RANDB.PPRD.Abbott.Com (SCHUBERT@CMDA) by RANDB.PPRD.Abbott.Com (PMDF V4.3-7 #5551) id <01HIXIZFRD3KEKO9Q5@RANDB.PPRD.Abbott.Com>; Mon, 31 Oct 1994 15:40:23 CDT Date: Mon, 31 Oct 1994 15:40:22 -0500 (CDT) Subject: G92-DFT/AlphaVMS To: chemistry@ccl.net Message-Id: <01HIXIZFTI9EEKO9Q5@RANDB.PPRD.Abbott.Com> X-Vms-To: RANDB::IN%"chemistry@ccl.net" X-Vms-Cc: SCHUBERT Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-Transfer-Encoding: 7BIT I am hoping someone can help with a couple of problems I am having during the testing of G92 on our Alpha running OpenVMS 6.1. One problem was solved by DEC. It required placing the following statement in MDUTIL.C before the "#include " statement: #define cma$tis_errno_get_addr CMA$TIS_ERRNO_GET_ADDR This cleared up a problem with symbol definition. Another problem remains in MDUTIL.C. The routine "isexe" does not correctly determine the "uid" and "gid" so G92 is not able to execute the required links. I got around this by setting the status (t = 1) in "getexen". The third problem is a bigger nuisance. When the archive values are printed at the end of a run by link L9999, the values for real numbers have been shifted by 2 bits, i.e., divided by 4. For example, if the output should read "HF = -72.4000", it will instead read "HF = -18.1000". Has anybody else run into these problems? Any help will be greatly appreciated. Jeff Schubert Abbott Laboratories schubert@cmda.abbott.com From criscuol@genny.gensia.com Mon Oct 31 17:42:46 1994 Received: from genny.gensia.com for criscuol@genny.gensia.com by www.ccl.net (8.6.9/930601.1506) id RAA08297; Mon, 31 Oct 1994 17:02:26 -0500 Received: by genny.gensia.com (931110.SGI/920502.SGI) for chemistry@ccl.net id AA17913; Mon, 31 Oct 94 13:52:57 -0800 Date: Mon, 31 Oct 1994 13:52:56 +48000 From: Fred Criscuolo To: chemistry@ccl.net Subject: Need source code for PDB library routines Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, I'm porting a raytracing application to the IBM RS/6000. Part of the code we received has an IRIX binary library of PDB routines. The routines are: rw-r--r--2401/3000 1164 Jun 30 11:16 1992 PDBLookup.o rw-r--r--2401/3000 1400 Jun 30 11:16 1992 PDBOffsets.o rw-r--r--2401/3000 2472 Jun 30 11:16 1992 PDBReadATab.o rw-r--r--2401/3000 6492 Jun 30 11:17 1992 PDBReadRec.o rw-r--r--2401/3000 1108 Jun 30 11:17 1992 PDBReadRes.o rw-r--r--2401/3000 1484 Jun 30 11:17 1992 PDBVersion.o rw-r--r--2401/3000 3356 Jun 30 11:17 1992 PDBWriteRec.o While this library, and the functions it supports, can be excluded from the make process, I'd like to incorporate them. I've scanned the public domain software available from Brookhaven but couldn't locate this library. If anyone recognizes these routines and can tell me where they're from I'd appreciate it. In the application I'm porting this library is called libpdb.a. Thanks in advance. Fred Fred Criscuolo Chemistry and Biochemistry (619) 622-4113 Gensia, Inc. criscuol@gensia.com O "There's more to life than rowing.. /--/ But not much." _________________________//|/_______________________ \__________________________/_______________________/ ~~~~~~~~~ San Diego Rowing Club / est. 1888 ~~~~~~~~~~~~~~~~~ / / From C1790@SLVAXA.UMSL.EDU Mon Oct 31 18:22:36 1994 Received: from SLVAXA for C1790@SLVAXA.UMSL.EDU by www.ccl.net (8.6.9/930601.1506) id RAA08677; Mon, 31 Oct 1994 17:24:08 -0500 Received: from SLVAXA.UMSL.EDU by SLVAXA.UMSL.EDU (PMDF V4.3-7 #2503) id <01HIXL2XTP1S95MLUO@SLVAXA.UMSL.EDU>; Mon, 31 Oct 1994 16:24:47 CDT Date: Mon, 31 Oct 1994 16:24:47 -0500 (CDT) From: BILL WELSH Subject: molecular modeling software for Sun SparcStations To: chemistry@ccl.net Message-id: <01HIXL2XZL8I95MLUO@SLVAXA.UMSL.EDU> X-VMS-To: in%"chemistry@ccl.net" MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT Dear Netters, From someone more familiar with SGI software, I'm looking for molecular modeling software that will run on a Sun SparcStation5 with a SparcII processor running the Solaris 1.1.1 O/S. Price is an issue (of course). I'm primarily interested in widely-distributed programs, commercial or otherwise (e.g., Insight, Cerius_2, Sybyl, Hyperchem, Macromodel, Semichem, Spartan, Gaussian [or other ab initio package]). Thanks ... Bill Welsh Dept. of Chemistry Univ. of Missouri-St. Louis St. Louis, MO 63121 From mmcallis@alchemy.chem.utoronto.ca Mon Oct 31 21:22:45 1994 Received: from alchemy.chem.utoronto.ca for mmcallis@alchemy.chem.utoronto.ca by www.ccl.net (8.6.9/930601.1506) id UAA11515; Mon, 31 Oct 1994 20:30:26 -0500 Received: (from mmcallis@localhost) by alchemy.chem.utoronto.ca (8.6.9/8.6.9) id UAA12613 for chemistry@ccl.net; Mon, 31 Oct 1994 20:30:21 -0500 Date: Mon, 31 Oct 1994 20:30:21 -0500 From: Mike McAllister Message-Id: <199411010130.UAA12613@alchemy.chem.utoronto.ca> To: chemistry@ccl.net Subject: PMP3 ?? Does anyone know how (simply) to get the Gaussian programs to calculate the PMP3 or PMP4 single point energies of ground state triplet wavefunctions? Thank you, Mike McAllister From o-cha@cnj.digex.net Mon Oct 31 21:28:27 1994 Received: from cnj.digex.net for o-cha@cnj.digex.net by www.ccl.net (8.6.9/930601.1506) id UAA12475; Mon, 31 Oct 1994 20:47:46 -0500 Received: by cnj.digex.net id AA25662 (5.67b8/IDA-1.5 for chemistry@ccl.net); Mon, 31 Oct 1994 20:47:17 -0500 From: "Matthew E. Harbowy" Message-Id: <199411010147.AA25662@cnj.digex.net> Subject: Re: CCL:Convergence in MOPAC C.I. calculations To: KEWISE@aardvark.ucs.uoknor.edu, chemistry@ccl.net Date: Mon, 31 Oct 1994 20:47:14 +1900 (EST) In-Reply-To: <199410311423.JAA28673@www.ccl.net> from "KEWISE@aardvark.ucs.uoknor.edu" at Oct 31, 94 08:24:00 am X-Mailer: ELM [version 2.4 PL24beta] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2154 >I am using MOPAC 6.0 to do calculations on fairly large polycyclic >molecules related to porphyrins. Part of this work involves optimizing >geometries and calculating energies for the cations and anions of these >molecules. To make these calculations more accurate (hopefully) we are >using the configuration interaction options in MOPAC 6, but have encountered >some problems with convergence. After reducing the gradient to below 1.0, It's a great idea, one that i've tried to rationalize myself. It's also a double correction, since parameterization adjusts the energies to include the 'correlation' piece once already, and what that means for transition states is anyone's guess. >the optimization begins to oscillate, and shows no sign of convergence >after running for several hours. A typical command line is as follows: > >AM1 PRECISE CHARGE=-1 C.I.=(4,2) VECTORS NODIIS Indeed, I saw the same thing, and EF only makes it worse. NLLSQ was the one that did the trick, and occasionally I would use SIGMA as well. Also, you have to be very careful about picking your ci levels. sometimes (5,2) is a much better choice due to near degeneracies which don't cause it to abort with errors, but oscillate instead. Also, try dummy atoms, and define your geometries in different ways. A well chosen zmatrix cured some, not all, of my ills. Part of my problem turned out to be I was attributing more symmetry to the molecule than was true by a bad zmatrix definition. XYZ was right out because I was studying a cyclization, and the orchestrated nature of the reaction required good control of the zmatrix. I assume for porphyrin analogs you're using symmetry to cut down on the number of dependent variables. what I found was the forced symmetry led me to a multiple saddle point, not a minimum or transition state. >Kristopher Wise >University of Oklahoma >Department of Chemistry and Biochemistry >KEWISE@aardvark.ucs.uoknor.edu matt harbowy tea chemist, thomas j lipton. -- matt harbowy | o-cha@cnj.digex.net | "I'm the bear that went greentea@eden.rutgers.edu | over the mountain." harbowy@aol.com |