From UDIM018@FRORS31.bitnet Fri Oct 19 08:23:31 1992 Date: 19 Oct 92 12:23:31 EDT From: Subject: e-mail address request To: chemistry@ccl.net E. M. EVLETH Dynamique des Interactions Moleculaires Universite Pierre et Marie Curie 4 Place Jussieu, Tour 22, Paris 75005 33-1-44-27-42-08 (work), 33 = France; 1 = Paris 33-1-45-48-67-20 (home) FAX 33-1-27-42-17 e-mail UDIM018 at FRORS31 I would appreciate the e-mail address of: L. A. Curtiss, Argonne National Labs, Chicago or one of the coauthors (H. V. Brand or L. E. Iton) of a recent article in JPC 1992 96, 7725. From J_BROWN@uvmvax.uvm.edu Mon Oct 19 03:40:00 1992 Date: Mon, 19 Oct 1992 08:40 EST From: J_BROWN@uvmvax.uvm.edu Subject: Answer to question about activation parameter calculations To: chemistry@ccl.net Yajun Zheng asks: > In a recent paper, A. Warshel reported that the activation barrier of > enzymatic reactions can be calculated from one rate constant with a universal > preexponential factor. Since I do not have any experience with with, could > someone tell me how reliable is this approach? In our research, we calculate activation parameters from variable temp NMR data. We use at least 3 data sets (more are prefered), linear regression, and the Eyring equation. We have determined that the calculated free energy of activation is more accurate than either the calculated enthalpy or entropy of activation due to statistical reasons. (See a good p-chem book for details on the problems with extrapolation in the Eyring method and the ways to get around it.) From hogue@mosquito.den.mmc.com Mon Oct 19 02:07:01 1992 Date: Mon, 19 Oct 92 08:07:01 MDT From: hogue@mosquito.den.mmc.com (Pat Hogue 1-2183) To: chemistry@ccl.net Subject: molybdenum I appologize but I need the molybdenum parameters listed below for INDO, not CNDO as previously requested. Thank you, Pat Hogue Fpr s, p, d: Slater exponent, core integral and beta integral. Slater-Condon parameters: f0 for ss, sp, pp, sd, pd, dd f2 for pp, pd, dd g1 for sp, pd f4; g2 and g3 From DSMITH@uoft02.utoledo.edu Fri Oct 19 06:20:03 1992 Date: 19 Oct 1992 10:20:03 -0400 (EDT) From: "DR. DOUGLAS A. SMITH, UNIVERSITY OF TOLEDO" Subject: J-T discussion on anonymous ftp To: chemistry@ccl.net The Jahn-Teller discussion is being kept on anonymous ftp at the University of Toledo. I will update this file once per day as long as the discussion is on-going. Here is how you get this file using ftp: ftp FTP.UTOLEDO.EDU (or 131.183.1.1) login: anonymous password: your_e-mail_address (please !). ftp> cd [.CHEMISTRY] ftp> get jahn-t.osc ftp> quit. Doug Smith From DSMITH@uoft02.utoledo.edu Fri Oct 19 06:12:34 1992 Date: 19 Oct 1992 10:12:34 -0400 (EDT) From: "DR. DOUGLAS A. SMITH, UNIVERSITY OF TOLEDO" Subject: more on Azepine J-T distortion To: chemistry@ccl.net I want to thank all of you who have responded, and then to take the discussion to the next level. My original posting included the following statements: >The singlet behaves as one might expect, i.e. it is aromatic and planar at >both the RHF and CAS(6,7) levels, where six pi electrons are distributed >amongst the three pi orbitals, three pi* orbitals and the empty nitrogen >p(z) orbital. The CAS calculations indicate that the ground state >configuration is approximately 88% of the total wavefunction. The geometry >is of C2v symmetry. > >The triplet behaves abnormally, or at least not as I would have expected. >It is planar, despite the seventh electron in conjugation. And the system >apparently is undergoing a Jahn-Teller distortion from C2v symmetry. (All >this is at the UHF level. We were unable to get CAS optimizations to >converge, although CAS single points at the UHF optimized geometry did.) >We think that we have located a low lying normal mode vibration which clearly >shows this distortion. In response to many of you who pointed out that something of C2v symmetry cannot undergo a Jahn-Teller distortion because there are no degenerate electronic states for a point group with only 1 dimensional irreps, let me note: 1) I only called the geometry C2v, i.e. there are geometric variables that are similar enough for me to describe the _singlet_ as having two perpendicular C2 planes. 2) We never located a triplet with C2v symmetry. Only the distorted triplet, in which the C2 plane corresponding to the plane of the molecule remains but the C2 plane perpendicular to the molecular plane (i.e. it bisects the nitrogen and the C-C bond opposite) disappears. 3) When we constrain the triplet to contain a C2 plane perpendicular to the molecular plane, we find a true minima (no negative frequencies). This molecule, upon removal of the constraints, gives only the distorted triplet. 4) The CAS(6,7)//UHF/3-21G results indicate that the unconstrained triplet wavefunction is 90% composed of the ground state triplet wavefunction. So my first question is: Can this still be a J-T distortion? How do I determine the actual point group of the unconstrained but undistorted triplet, in order to see if this is really a J-T effect? My second question arises from a comment by Dave Bernholdt of QTP, who suggested that the results might be better classified as a pseudo- or second order J-T effect, in which low-lying states which can couple to the ground state via a suitable vibration do. So, I do I check for this? My third question arises from a posting by brl@inel.gov (sorry, but you did not sign your name), who said that the observed distortion is due to the "doublet instability problem" or "symmetry breaking." Well, the latter term seems obvious, but the first does not, or at least I am not understanding the terminology. Can someone expand on this comment? Thanks again for help, suggestions, and just the discussion. I plan to post the entire discussion via anonymous ftp - I will set that up today and send another mail message with the details, if any one is interested. Doug Smith Assistant Professor of Chemistry The University of Toledo Toledo, OH 43606-3390 voice 419-537-2116 fax 419-537-4033 email dsmith@uoft02.utoledo.edu From schinke@biochemtech.uni-halle.dbp.de Mon Oct 19 18:24:19 1992 Date: Mon, 19 Oct 1992 17:24:19 +0100 From: schinke@biochemtech.uni-halle.dbp.de To: chemistry@oscsunb.ccl.net Subject: Please, help with decision making! Dear netters, we intend to buy a new molecular modeling system for workstations (SGI). Other recommended us to buy either Insight/Discover from Biosym or Quanta >from Polygen Co. Up to now we are using SYBYL to solve our problems. Has anyone experiences with all three programs? Which is the best to investigate biomolecules from small peptides to enzyme-substrate interactions and to predict secondary structures of proteins. Or do you know another program which is best suited for these purposes. Are there known limitations or intrinsic bugs of these programs? It would be very helpful for use to know your opinion. Thanks a lot, Heiko Schinke =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= =-= Heiko Schinke =-= =-= Martin-Luther-University, Dept. of Biochemistry/Biotechnology =-= =-= P. O. Box 8, D-4020 Halle/Saale, Germany =-= =-= =-= =-= Phone Number: (0345) 617-242 FAX Number: (0345) 647-616 =-= =-= =-= =-= E-Mail: schinke@biochemtech.uni-halle.dbp.de =-= =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= From jan@si.fi.ameslab.gov Mon Oct 19 10:22:29 1992 From: jan@si.fi.ameslab.gov (Jan Jensen) Subject: Large Semiempirical calculations To: chemistry@ccl.net Date: Mon, 19 Oct 92 15:22:29 CDT Fellow netters, Here are some questions to all you semiempirical folks. What is the practical size limit on molecules one can do with semiempirical theory (e.g. AM1)? Is there a (un)official record for the largest system? It would seem that as computers get faster and memory and disk cheaper, quantum mechanical calculations should make their way into territory previously dominated by molecular mechanics (e.g. peptides, enzyme active sites, natural products, etc.). How far away are we from 500 or 1000 atom systems? Any tips, success stories, timings, etc. will be appreciated; and summarized for the list if there is sufficient interest. Jan Jensen Department of Chemistry Iowa State University jan@si.fi.ameslab.gov From jkl@ccl.net Mon Oct 19 12:51:16 1992 From: jkl@ccl.net (Jan Labanowski) Date: Mon, 19 Oct 1992 16:51:16 -0400 To: chemistry@ccl.net Subject: Results of vote Dear Subscribers, So now the results of vote on the future of Computational Chemistry List. There were 2 choices: A) I want the list to be moved from Ohio Supercomputer Center. B) I want the list to stay at the Ohio Supercomputer Center under current management. Unfortunately, we did not get a quorum (500) responses, what means that the majority of the subscribers are undecided (i.e., do not care what will happen to the list). Since you will ask anyhow, I will present the results which I do not consider binding. Now, how I counted the votes. I was callecting A and B votes in separate mail folders. Then I ran the following perl script on the A and B folders: ------------------ beginning of perl script #!/usr/local/bin/perl $VOTEFILE = $ARGV[0]; #take name of the file votes from command line #open file with votes open(VOTEFILE, "<$VOTEFILE") || die "Could not find $VOTEFILE\n"; #count votes $n = 0; while() { tr/A-Z/a-z/; if(/^from\s+(\S+)/) { $addresses[$n++] = $1; } } @sorted = sort @addresses; #sort votes by address printf STDOUT "Number of votes: %d\n", $n; #check for repeated addresses for ($i = 1; $i < $n; $i++) { if($sorted[$i] eq $sorted[$i-1]) { printf STDOUT "Address: %s repeated\n", $sorted[$i]; } } print STDOUT "All votes \n"; #write sorted addresses to stdout for ($i = 0; $i < $n; $i++) { printf STDOUT "%3d %s\n", $i+1, $sorted[$i]; } --------------------- end script ----- I had to massage the file, since there where some apparently repeated votes (e.g., from nobody@kodak.com, mail, etc.) and I had to correct the From line. Now the results: A: 8 B: 415 X: 1 (I presume X is: do not care) I have got also many comments on how to make the list better. I will take it into account and when I have more time, I will try to implement the changes which are unquestionable imporvements. The suggestions include: 1. Making the chemistry@ccl.net a News Group but preserving it also as an e-mail reflector (many newsgroups do that). You will have a choice to do one or the other or both. 2. Puting a GOPHER or WAIS server interface to our archives. Now. I want to do this, but I am held by two things: 1) I want to put the chemistry list, ftp site, etc. on a separate, stand-alone workstation. You might guess that Ohio Supercomputer Center is affected by the last few years of "booming economy". But there is a light in the tunnel. If somebody wants to advise me how to get money for a better support of this forum, I am all ears. Please understand, that it cannot be some cheap hardware, since the list distributes sometimes in excess of 30,000 messages a day. Try to do it to your workstation and see what happens. 2) There are several deadlines which I have to meet (yes, I also work...) and I simply do not have time to do much before the end of November. Now, for some of you the list internal working is transparent. For me it is not, and actually, there are many small improvements which were recently implemented in the script which distributes the mail: You may notice that you see less subscription messages on the list (actually I do not remember a single one in last few months). You may have noticed that all messages have Subject: lines. I also implemented an exclusion list. List now also sends a confirmation to people who post the message (except to mailers who return the "mail" address on the first line of header). You may say, these are all easy things --- maybe, I wonder how many of you really know how to do it. I did not and had to find out the hard way. There are also some other small improvements on its way. Finaly, some information about the list itself for those who refuse to read the help file for the list. You really should. The list, beside sending you messages, has archives. They hold all posted messages, but also have some reviews, free software and job clearing house. You may retrieve the following files for more information about the list: help --- rules and description of list services index --- index of files in the archives. The Subject lines of messages are also included software --- list of free software available from the archives positions.offered --- list of job opennings submitted to the list positions.wanted --- CV's of people looking for a job You can get these files either via e-mail by sending a message: send file-name from chemistry to OSCPOST@ccl.net or OSCPOST@OHSTPY.BITNET (file-name is one of the files described above) or by ftp to www.ccl.net [128.146.36.48], e.g.: ftp www.ccl.net (or ftp 128.146.36.48) Login: anonymous Password: Your_email_address ftp> ascii ftp> cd pub/chemistry ftp> get file-name ftp> quit Sorry for taking so much bandwidth with administration. Jan Labanowski Ohio Supercomputer Center jkl@ccl.net From d3f012@gator.pnl.gov Mon Oct 19 08:32:04 1992 Date: Mon, 19 Oct 92 15:32:04 PDT From: d3f012@gator.pnl.gov Subject: Re: Large Semiempirical calculations To: chemistry@ccl.net, jan@si.fi.ameslab.gov > Fellow netters, > > Here are some questions to all you semiempirical folks. What is the > practical size limit on molecules one can do with semiempirical theory > (e.g. AM1)? Is there a (un)official record for the largest system? > It would seem that as computers get faster and memory and disk cheaper, > quantum mechanical calculations should make their way into territory previously > dominated by molecular mechanics (e.g. peptides, enzyme active sites, > natural products, etc.). How far away are we from 500 or 1000 atom > systems? Any tips, success stories, timings, etc. will be appreciated; > and summarized for the list if there is sufficient interest. > > Jan Jensen > Department of Chemistry > Iowa State University > jan@si.fi.ameslab.gov > Jan, 500 atom systems have been "here" for at least a couple of years. I published results of semiempirical UV/vis spectroscopic calculations using INDO/s on the chromophores of the photosynthetic reaction center of Rps. viridis. [JACS 113, (1991) p. 8210] These calculations contained: 536 atoms, 1532 electrons,1436 basis functions. The single-excited CI used the lowest 1680 single excited configurations from the ground state. These size calculations are very routine. At the time, they were done on a Cray YMP. Now, they can be done on IBM 550-strength workstations. I have calculations in preparation that are about 3-4 times larger. Access to a Cray C90 will be very helpful for these. We are also routinely running semi-empirical MD calculations (with forces >from Hartree Fock) on 60-100 atom systems on workstations. Mark Thompson ************************************************************************** Mark A. Thompson email: d3f012@pnlg.pnl.gov Molecular Science Research Center FAX : 509-375-6631 Pacific Northwest Laboratory voice: 509-375-6734 PO Box 999, Mail Stop K1-90 Richland, WA. 99352 ************************************************************************** From m10!frisch@uunet.UU.NET Mon Oct 19 14:57:11 1992 Date: Mon, 19 Oct 92 18:57:11 EDT From: m10!frisch@uunet.UU.NET (Michael Frisch) Subject: Re: Large Semiempirical calculations To: chemistry@ccl.net Fellow netters, Here are some questions to all you semiempirical folks. What is the practical size limit on molecules one can do with semiempirical theory (e.g. AM1)? Is there a (un)official record for the largest system? It would seem that as computers get faster and memory and disk cheaper, quantum mechanical calculations should make their way into territory previously dominated by molecular mechanics (e.g. peptides, enzyme active sites, natural products, etc.). How far away are we from 500 or 1000 atom systems? Any tips, success stories, timings, etc. will be appreciated; and summarized for the list if there is sufficient interest. Jan Jensen The DNA fragment in the Gaussian 92 poster is 378 atoms, and a HF/STO-3G single point only takes about 2 days on a workstation (RS/6000-550). So 500 atoms is thinking pretty small for a semiempirical calculation, at least for a single point. I know that people at Cray have done over 500 atoms in MNDO90, but I'm don't think they worked up much of a sweat doing something that size. An optimization is much more difficult, of course, because of the large number of degrees of freedom (to say nothing of the multiple minima problem). Mike Frisch -------