From m10!frisch@uunet.UU.NET Wed Sep 9 18:53:32 1992 Date: Wed, 9 Sep 92 22:53:32 EDT From: m10!frisch@uunet.UU.NET (Michael Frisch) To: uunet!csv.warwick.ac.uk!msrag@uunet.UU.NET Subject: Gaussian90/MP2 Analytic Force Constants? Dr. Buttar asked about MP2 frequencies in Gaussian 9x. Gaussian 90 did MP2 frequencies by numerically differentiating the analytic first dervatives. Gaussian 92 does MP2 frequencies using analytical second derivatives. For a molecule with a fair number of atoms this is significantly faster. For example, 6-31G* pentanol, C1 symmetry, is about 10 times faster in G92 than G90. Mike Frisch From wipke@SECS.UCSC.EDU Wed Sep 9 16:35:19 1992 Date: Wed, 09 Sep 1992 23:35:19 PDT From: "W. Todd Wipke" To: chemistry@ccl.net Subject: Re: bibtex style files for chemistry journals Mingzuo Shen inquired if these exist. The TCM style is the current Tetrahedron style and is nearly identical to JACS and JCICS style. You can obtain this from TCM vol 3:5 disk 13, or send msg to "fileserv@secs.ucsc.edu and in the body put one line SENDME INSTR.TEX* This supports all the bib types used by my REFFORM program. REFFORM v1.1 is on disk 12 of same issue, and enables MS word to do the same things for bib. -Todd Wipke UCSC From jstewart@fai.fai.com Wed Sep 9 14:69:21 1992 Date: Thu, 10 Sep 92 08:04:21-1795 From: Dr. James Stewart To: chemistry@ccl.net Subject: Some Semiempirical Thoughts Some Semiempirical Thoughts 1. As Joe Leonard correctly suggested, I am not an employeee of Fujitsu. Stewart Computational Chemistry is a private consulting concern. Fujitsu is a customer. 2. I regret my observation on SAM-1. However, let me explain. The results I reported for SAM-1 were taken from an official publication written by Prof Dewar, and approved for public release. 3. Drs Caoxian Jie and Jianguo Yu expressed interest over the Gaussians in AM1. The history of these terms is as follows: In 1982 Prof Dewar asked me to work on correcting the known error in MNDO, specifically to correct for the excessive repulsions at roughly the van der Waals' distances. After a short time, it became clear that the fault could be corrected using the method of operator equivalents: a modification to the core-core terms could mimic the true correction, which should more correctly have been made elsewhere. A Gaussian series is like a Fourier series: you can model any function you want with it. For that reason, I chose to add Gaussians to the core-core terms. This was intended as an interim measure; at least, that was my thought at that time. Hopefully, we can soon do away with operator equivalents, which really only treat the symptom and not the cause. 4. MOPAC-7 will be coming out in a few months. Details are still not finalized, but should be in about a month. At the present time, however, it looks likely that MOPAC-7 will have the following features: e. Most important, the amount of high-quality code contributed by various people is at an alltime high. d. Some symmetry has been added. b. One, two, and three dimensional systems will be supported. c. Electronic band-structures will be calculable. f. In one form, MOPAC-7 will be placed in the public domain. g. The Manual has been expanded. This is in response to so many requests for more detail. a. Copies of the program and the manual will be available from several sources, including QCPE. None of this is definite, of course. For just now, MOPAC-7 should be regarded as `vaporware'. James J. P. Stewart Tel 719 488 9416 FAX 719 488 9758 E-mail jstewart@fai.com From ckf@f16.cray.com Thu Sep 10 08:10:39 1992 Date: Thu, 10 Sep 92 12:10:39 EDT From: ckf@f16.cray.com (Charles Foley) To: chemistry@ccl.net Subject: Allen & Tildesley code by ftp? I tried getting the source code that was mentioned in Allen & Tildesley's book "Computer Simulations of Liquids" by the anonymous ftp procedure outlined in Appendix F. It seems that the account may have been disabled because I couldn't get in using the account specified in the appendix or by anonymous ftp. Does anyone know of an alternative site for this code, or, if anyone has the code, would you consider e-mailing it to me? Thanks for your assistance, Charles K. Foley, Ph.D. Cray Research, Inc. Americas & Asia/Pacific Group P.O. Box 12746 Research Triangle Park, NC 27709 o <( ckf@f16.cray.com [\. From berkley@wubs.wustl.edu Thu Sep 10 08:57:13 1992 Date: Thu, 10 Sep 92 13:57:13 -0500 From: berkley@wubs.wustl.edu (Berkley Shands) To: chemistry@ccl.net Subject: Receptor Timings Again Receptor (TM - Tripos Associates) Beta version V2.3 timings for the ACE series at four and two degree scans. The first molecule in the series is five rotatable bonds. Runtimes reported for V2.3 are truncated to the nearest integer second for compatibility with V2.2. The 4 and 2 degree scans converge to one cluster of two points. A post-processed run has 14 points in two clusters. A run with an automatically generated, variable scan parameter ranging from 2.8 to 0.86 degrees yeilds 5 points in two clusters without post-processing. ******************************************************************************* Benchmarks of the complete ACE series of 71 molecules at four degrees ******************************************************************************* Top CPU Top Elps Top Sys Chg CPU Chg Elps CPU Type and notes ======= ======== ======= ======= ======== ================== XXXX XXXX XXXX 181 XXXX IRIS Indigo (IP12) Receptor V2.2 741 772 198 484 688 VS3520 32mb VMS 5.5-2+VIP 120 148 10 67 83 SUN 4mp Sparc 690 64Mb 4-cpu system 125 143 4 68 86 SUN 4/c SPARCstation 2 28mb 83 153 11 55 95 ESV M120 R3000A@25Mhz 100 138 2 46 64 IBM RS-6000-320H 32mb (AIX 3.2.1) 65 86 11 40 50 IRIS Indigo (IP12) 48mb R3000A@35Mhz 71 91 3 43 57 IRIS 4d/380 (IP7) 128mb R3000A@35Mhz 63 187 3 34 91 IBM RS-6000-550 64mb (AIX 3.2.1) 49 64 1 23 32 IBM RS-6000-560 96mb (AIX 3.2.1) ******************************************************************************* Benchmarks of the complete ACE series of 71 molecules at two degrees Theoretical conformations counts range from 10^6 to 10^20 at two degrees. (Don't try this with Systematic Search, folks) ******************************************************************************* Top CPU Top Elps Top Sys Chg CPU Chg Elps CPU Type and notes ======= ======== ======= ======= ======== ================== XXXX XXXX XXXX 8905 XXXX IRIS Indigo (IP12) Receptor V2.2 XXX 3968 XXX 5721 6066 VS3520 32mb VMS 5.5-2+VIP 731 776 6 669 706 SUN 4/c SPARCstation 2 28mb 661 713 5 625 654 ESV M120 R3000A@25Mhz 631 683 2 573 603 IBM RS-6000-320H 32mb (AIX 3.2.1) 512 346 27 487 514 IRIS 4d/380S (IP7) 128mb 3/8 CPUs 502 568 24 475 526 IRIS Indigo (IP12) 48mb R3000A@35Mhz 356 378 2 323 334 IBM RS-6000-550 64mb (AIX 3.2.1) 300 317 1 268 286 IBM RS-6000-560 96mb (AIX 3.2.1) ******************************************************************************* The ace-28 series ran 84,413 seconds with V2.2 and 8,718 with V2.3 at 2 degrees. ******************************************************************************* Notes: "Top CPU" is the recorded total CPU for all forked processes, including overhead and I/O. "Top Elps" is the recorded elapsed time from command line parsing to the exit of the last child processes. "Top Sys" is the recorded UNIX system overhead from the times() function "Chg CPU" is the algorithmic chargable time from initial rotations to termination (child process CPU) "Chg Elps" is the elapsed time spent rotating and computing. From ravishan@swan.wesleyan.edu Thu Sep 10 10:42:51 1992 Date: Thu, 10 Sep 92 14:42:51 -0400 From: G. Ravishanker To: chemistry@ccl.net Subject: C. Altona Hi If anyone knows how C. Altona's (Leiden, The Netherlands) e-mail address, I would appreciate hearing from you. I have to clarify some things >from a paper of his. Thanks in advance. Ravi From jle@world.std.com Thu Sep 10 17:18:22 1992 Date: Thu, 10 Sep 1992 21:18:22 -0400 From: jle@world.std.com (Joe M Leonard) To: chemistry@ccl.net Subject: Anomeric effect in MM3 A question for those in the know... Is the anomeric effect in MM3 included in the gradient and hesian calculations or is it "added on" after each iteration. Since the various literature sources review the results rather than the implementation of MM3, it is not obvious from reading them... Also, JACS 114, 6120 refers to a new form of the H-bond term - is the angle omega defined as X-H ... O-X -or- X-H ... O-H or...? If it's a torsional angle, what is the "fourth" atom? Finally, the same paper referred to hierarchichal (sp?) types, in that complex types revert to general types when lacking parameters. Is the cascade what would be obvious, or are there "can't become" and "must become" options that are not described in the paper? Replies to me can save net bandwidth, although I LOVE good discussions on the net! Joe Leonard jle@world.std.com