From m10!frisch@uunet.UU.NET Sun Sep 20 19:50:40 1992 Date: Sun, 20 Sep 92 23:50:40 EDT From: m10!frisch@uunet.UU.NET (Michael Frisch) Subject: Re: GAUSSIAN Workshops To: chemistry@ccl.net, uunet!iscsvax.uni.edu!MACMILLAN@uunet.UU.NET Can anyone tell me when and where there will be any workshops on GAUSSIAN in the midwest? We have purchased GAUSSIAN 92 for our RS6000 and some faculty would like to attend a workshop on how to use it. You can respond directly to me. Jim Macmillan University of Northern Iowa MACMILLAN@ISCSVAX.UNI.EDU We're in the process of arranging locations for next year's workshops. We'll post an announcement to this mailing list as soon as the location(s) and date(s) are settled. Mike Frisch ------- From m10!frisch@uunet.UU.NET Sun Sep 20 19:39:06 1992 Date: Sun, 20 Sep 92 23:39:06 EDT From: m10!frisch@uunet.UU.NET (Michael Frisch) Subject: Re: Tell me which is right, G90 or MOPAC? To: chemistry@ccl.net uunet!helix.nih.gov!milan (Milan Hodoscek) asks: Hi, Comparing results from different programs I encountered this problem: C6H6 (kcal/mole) MNDO AM1 MOPAC 21.87985 23.06206 G90 21.76903 22.96633 1. Note that runs of the same MOPAC code on different systems (different sompilers and hardware) can vary by this much if the default accuracy level is used. 2. Gaussian defaults to PRECISE, so to make comparsions with MOPAC, this keyword should be used. This is probably the source of your discrepancies. Mike Frisch ------- From cabku01@mailserv.zdv.uni-tuebingen.de Mon Sep 21 14:18:02 1992 Date: Mon, 21 Sep 92 12:18:02 +0200 From: cabku01@mailserv.zdv.uni-tuebingen.de (Hartwig Kuehbauch) To: chemistry@ccl.net Subject: 2nd RFD: sci.chem.organomet Dear netters! This is the second request for discussion regarding creation of the newsgroup sci.chem.organomet A summary of arguments after the first RFD: Most people who e-mailed me appreciated the idea of creating a new newsgroup for organometallic chemistry very much! Those who did not, normally didn't have objections on the newsgroup itself, but they said, that it would be better, if there was a newsgroup called sci.chem.organic with a subsection for organometallic chemistry, and that, if I raise up a sci.chem.organomet newsgroup I also have to raise up other newsgroups like sci.chem.organic, sci.chem.inorganic, and so on... Well that's easy to answer. Organometallic chemistry uses some tools of the organic chemists, but it is something completely different. Chemistry is such a widespread area, that nobody can really feel comfortable in only one newsgroup. It's true, there should be newsgroups like sci.chem.organic, and I know a lot of people who would appreciate the creation of such newsgroups very much (me too !!!), but I'm a little bit oldfashioned. One after the other! I am working on organometallics, not on organic chemistry. :-) There also was the objection, that normally a new newsgroup only makes sense, if the "mother-newsgroup" (here sci.chem) is inundated with articles (which sci.chem is not!). Well, to say it in economic terms: if there is a demand, there will be a supply for that demand --- and vice versa! Most of the chemists, that e-mailed me after the first RFD, told me, that they rarely read sci.chem, because they don't want to waste too much time browsing through articles which are of no interest for them. But they would read regularly a newsgroup called sci.chem.organomet!!! It's true diversifica- tion doesn't always make sense, but sometimes it's unavoidable! And I'm sure here it is necessary! Sci.chem is too common and unattractive for most people. After all, creating a new newsgroup sci.chem.organomet wouldn't bring up a single disadvantage --- so why shouldn't we try it? Ok, that's all for today. Greetings from Germany. Hartwig Kuehbauch ============================================================================= = Hartwig Kuehbauch - University of Tuebingen - Dep. of Inorg. Chemistry II = = (cabku01@mailserv.zdv.uni-tuebingen.de) - Germany - = ============================================================================= From UDIM018@FRORS31.bitnet Fri Sep 21 11:02:10 1992 Date: 21 Sep 92 15:02:10 EDT From: Subject: meeting To: chemistry@ccl.net E. M. EVLETH Dynamique des Interactions Moleculaires Universite Pierre et Marie Curie 4 Place Jussieu, Tour 22, Paris 75005 (1) 44 27 42 08 UDIM018 at FRORS31 *********************************************************************** SYMPOSIUM ON THE DESIGN OF MOLECULES AND COMPUTATIONAL CHEMISTRY and the IBEROAMERICAN WORKSHOP ON ENVIRONMENTAL ANALYTICAL CHEMISTRY *********************************************************************** The sympsium on the design of molecules and computational chemistry (with the participation from the Universities of Uppsala and Stolkhom) and the Iberoamerican Workshop on Environmental Analytical Chemistry will be held on 2-5 February 1993 at the University of Oriente, Santiago de Cuba as part of the 14th Conference on Chemistry. The conference is jointed organized by the University of Oriente, the Cuban Chemical Society and the Cuban Academy of Sciences. The themes. Organic Chemistry Mining Inorganic Chemistry Zeolites Analytical Chemistry Chemistry and human healt Physical Chemistry Biochemistry Chemistry of Sugar Cane Derivatives Treatment and uses of wastes Natural Products X-ray crystallogaphy Spectroscopy and Structure For further information write to Professor Marieta Gomez Serrano, XIV Conference on Chemistry, Santiago de Cuba 90500 Cuba before October 4, 1992. ******************************************************************* US citizens should check with the US Treasury Department to see if you can to entry Cuba for this meeting (if you wish to attend). My reading of document furnished by the US Embassy in Paris indicates that you do, this meeting does not fall under the embargo. But some people may have problems, especially for their spouses. ******************************************************************** From system@alchemy.chem.utoronto.ca Mon Sep 21 07:43:22 1992 Date: Mon, 21 Sep 92 11:43:22 -0400 From: system@alchemy.chem.utoronto.ca (System Admin (Mike Peterson)) To: chemistry@ccl.net Subject: Re: Tell me which is right, G90 or MOPAC? I have run the C6H6 test case on MOPAC 6 on our HP 750, and I get the same result as Milan Hodoscek for both MNDO and AM1. Using PRECISE only changes the energy in the fifth decimal place (I have seen PRECISE make a lot more difference if the geometry is optimized, which it is not in this case). Mike. -- Mike Peterson, System Administrator, U/Toronto Department of Chemistry E-mail: system@alchemy.chem.utoronto.ca Tel: (416) 978-7094 Fax: (416) 978-8775 From NEELY@AUDUCVAX.bitnet Mon Sep 21 06:14:00 1992 Date: Mon, 21 Sep 92 11:14 CDT From: Subject: FORCE & UHF with AMPAC/MOPAC To: chemistry@ccl.net I've been trying to locate a transition state in a system with unpaired electrons using AMPAC/MOPAC. The ROHF computations don't converge for me, but it's not surprising, as I have a lit- erature reference for a similar system that had to be treated using UHF. The UHF computations converge, but when I try to do a FORCE computation on the result, both AMPAC & MOPAC 'opti- mize' this system to product before computing the force constants. It's not clear to me from the manuals that the FORCE & UHF keywords are incompatible. Any advice would be welcome. Thanks! Irene Newhouse From d3f012@gator.pnl.gov Mon Sep 21 02:56:28 1992 Date: Mon, 21 Sep 92 09:56:28 PDT From: d3f012@gator.pnl.gov Subject: Re: transition dipole moments and magnetic moments To: chemistry@ccl.net, lrellick@magnus.acs.ohio-state.edu > From chemistry-request@ccl.net Fri Sep 18 10:26:30 1992 > Resent-From: chemistry-request@ccl.net > Resent-Date: Fri, 18 Sep 92 10:01 PDT > Subject: transition dipole moments and magnetic moments > Sender: chemistry-request@ccl.net > Resent-To: d3f012@gator.pnl.gov, D3E129@CAGLE.PNL.GOV, d3e057@bogart.pnl.gov, > D3E353@MINERVA.PNL.GOV > To: chemistry@ccl.net > Resent-Message-Id: > X-Envelope-To: d3f012@gator.pnl.gov > X-Vms-To: chemistry@ccl.net > Content-Length: 499 > X-Lines: 12 > > Dear Netters, does anyone know which software program I can use to calculate > the transition dipole moments and magnetic moments for molecules containing > 20-30 atoms (approx.)? Any (ANY) help will be appreciated! > > - Lori Rellick, Biochemistry Dept. OSU > > --- > Administrivia: This message is automatically appended by the mail exploder: > CHEMISTRY@ccl.net: --- everyone CHEMISTRY-REQUEST@ccl.net: --- coordinator > OSCPOST@ccl.net: send help from chemistry Anon. ftp: www.ccl.net > --- > > Lori, Argus can calculate the quantities you are interested in (at an INDO/S level). Transition dipole moments support both the length and velocity form of the dipole moment operator, while magnetic moments use the dipole velocity based operator. Some good references on these formalisms: 1. Bouman and Hansen J. Chem. Phys. 66 (1977), 3460. 2. Hansen and Bouman Molecular Physics 37 (1979), 1713. 3. Hansen and Bouman Chem. Phys Letts 45, (1977), 326. The program is available, via anonymous ftp, from pnlg.pnl.gov (130.20.64.11) in the argus subdirectory. Hope this helps. Mark Thompson d3f012@pnlg.pnl.gov From wsonnen@alnitak.usc.edu Mon Sep 21 03:25:52 1992 Date: Mon, 21 Sep 92 10:25:52 PDT From: wsonnen@alnitak.usc.edu (Wayne Sonnen) To: chemistry@ccl.net Subject: calc of dielectric const. Monica, with respect to your question about calculation of dielectric constants using MD.... It is not clear to me at what level you wish to calculate dielectric constant (i mean dielectric permittivity). From an MD calculation, i see two levels: with and w/o polarizabilities in the water model. Regardless of the water model, you can use equation 5.25 in Allen and Tildesley's book. Page 163, second last paragraph, states the importance of the external relative permittivity. There should be sufficient references in this book. Additional references on water models (i have lots of them) are: (1) H.J.C. Berendsen, JPC (1987) {\bf 91}, 6269-6271 (2) W.L. Jorgensen, JCP (1983) {\bf 79}, 926-935 (3) U. Niesar, ACS (1990) {\bf 94}, 7949-7956 (4) P. Ahlstrom, Mol Phys (1989), {\bf 68} 563-581 (5) A. Rahman, JCP (1971), {\bf 55}, 3336-3359 (6) D. L. Beveridge, In: 'Molecular-Based Study of Fluids', Chapter 13. (1982 ??). (sorry, i don't have a complete ref.; editor ??). Ref. #(5), equation 4.15 discusses the calculation of 'dielectric constant' >from MD simulation. The equation may appear to be complicated, but this particular equation simply uses the static dipole moment of water (about 1.85 D) and two calculated quantities from the simulation: the polarizability and the (permanent dipole) orientational correlation function. good luck, Wayne Sonnen From lrellick@magnus.acs.ohio-state.edu Mon Sep 21 11:22:30 1992 Date: Mon, 21 Sep 92 15:22:30 EDT From: Lorraine M Rellick To: chemistry@ccl.net Subject: partial charges - g92 Dear Netters, does anyone know how to output partial charges from G92 (I am using CIS as the method). - Thanks!, Lori Rellick Biochem Dept, OSU From milan@helix.nih.gov Mon Sep 21 11:33:33 1992 Date: Mon, 21 Sep 92 15:33:33 -0400 From: milan@helix.nih.gov (Milan Hodoscek) To: CHEMISTRY@ccl.net, milan@helix.nih.gov Subject: More details on MOPAC/G90 problem. Dear netters, I want to put some more details about the problem of different results reported from G90 and MOPAC for the same systems. 1. This error is present in all systems not only benzene; I started with water but since MOPAC cannot use cartesian coordinates on triatomic molecules I switched to C6H6. 2. To avoid rundoff errors in converting coordinate systems I used cartesian coordinates instead of internal ones. 3. Let me extend the table I gave before a little further: C6H6 (kcal/mole) on Convex MNDO MNDO(precise) AM1 MOPAC 21.87985 21.87987 23.06206 G90 21.76903 22.96633 tokcal(mopac) 630.7040 630.1250 C6H6 (kcal/mole) on Cray MNDO MNDO(prec) MOPAC 21.87984 21.87987 G90 21.76919 TSB(C14H12) MNDO MOPAC 75.90669 G90 75.65258 4. The above tables show that results don't vary so much on different hardware. The same is true for the PRECISE keyword. 5. The situation get worse if we deal with somewhat bigger systems. Difference is already one quarter of kcal on 36 atom system. 6. This isn't a conversion constant problem: conversion factor in G90 is 1 Hartree = 627.5095 kcal/mole. (tokcal=627.5095) If we calculate what this constant would be in MOPAC to give the same result as G90 than tokcal = 630.7040 for MNDO and 630.1250 for AM1 method in the same program, which is very unlikely. 7. Some people reported the same as the above MOPAC 6.0 numbers on the following hardware and/or software: HP/750 MOPAC 6.0 SG/4Dgt70 AMPAC 3.01 Unknown AMPAC 4.0, AMPAC 2.1 8. I changed also position of first atom in benzene for 0.001 A and the difference in energy was only 0.00196 kcal/mole. The difference in porgrams is 0.01082. I consider 0.001 A significant change in geometry in terms of roundoff errors. Conclusion: It seems to me that this isn't roundoff error nor conversion factor problem, but either of the two: 1. parameters for C and/or H atoms are different 2. code is not implemented correctly Regards -- Milan Hodoscek From NETNEWS@UICVM.UIC.EDU Mon Sep 21 17:52:38 1992 To: CHEMISTRY@ccl.net Organization: University of Illinois at Chicago Date: Monday, 21 Sep 1992 14:21:42 CDT From: Subject: MD of macromolecules (Summary) Folks: Here is a list of books/articles/etc I obtained in response to my posting requesting references on MD of Polymers. The best bet would be the first book on the list which has a few review papers and also a whole bunch of references. Sowmyan ---------------------------------------------------------------------------- The recent book, Computer Simulation of Polymers, ed. R.J. Roe, Prentice Hall, 1991 is a possible source. About half of the 28 chapters deal with various applications of dynamics to polymer problems. Kathy Clark kathy@tripos.com -------------------------------------------------------------------------- One book you should certainly take a look at is "Dynamics of Proteins and Nucleic Acids" by McCammon and Harvey, published by Cambridge University Press in 1987. --------------------------------------------------------------------------- Osiris Consultants has recently been formed, and is dedicated to serving the needs of the chemistry community. While many scientists are taking advantage of computational chemistry techniques, there are many others who are not sure how these techniques can be applied to their particular problems. Osiris Consultants has been formed to serve both groups. Osiris Consultants is available as a source of information about, and assistance with, problems in all areas of computational chemistry, especially the areas of computer-aided drug design and protein modeling. We can provide guidance on how to approach a particular research problem, or conduct computational research. We are available for assistance with general inquiries >from the basic to the very specific. Custom Sybyl Programming Language (SPL) progamming also available. For more information, contact Lisa M. Balbes, Ph.D., at 614-279-1829, or write to us at 1371 Tradewinds Dr,. Suite 202, Columbus, OH 43204. Electronic mail may be sent to balbes@osiris.rti.org. ------------------------------------------------------------------------------- Article{Gunsteren90, author = {Wilfried F. van Gunsteren and Herman J. C. Berendsen}, title = {Computer Simulation of Molecular Dynamics: {M}ethodolo gy, Applications, and Perspectives in Chemistry}, journal = ACI, volume = 29, pages = {992-1023}, year = 1990, annote = {review}, } From d3f012@gator.pnl.gov Mon Sep 21 10:08:05 1992 Date: Mon, 21 Sep 92 17:08:05 PDT From: d3f012@gator.pnl.gov Subject: Re: More details on MOPAC/G90 problem. To: CHEMISTRY@ccl.net, milan@helix.nih.gov > > Dear netters, > > I want to put some more details about the problem of different results > reported from G90 and MOPAC for the same systems. > > 1. This error is present in all systems not only benzene; I started with > water but since MOPAC cannot use cartesian coordinates on triatomic > molecules I switched to C6H6. > > 2. To avoid rundoff errors in converting coordinate systems I used > cartesian coordinates instead of internal ones. > > 3. Let me extend the table I gave before a little further: > > C6H6 (kcal/mole) on Convex > > MNDO MNDO(precise) AM1 > MOPAC 21.87985 21.87987 23.06206 > G90 21.76903 22.96633 > tokcal(mopac) 630.7040 630.1250 > > C6H6 (kcal/mole) on Cray > > MNDO MNDO(prec) > MOPAC 21.87984 21.87987 > G90 21.76919 > > > TSB(C14H12) > MNDO > MOPAC 75.90669 > G90 75.65258 > > 4. The above tables show that results don't vary so much on different > hardware. The same is true for the PRECISE keyword. > > 5. The situation get worse if we deal with somewhat bigger systems. > Difference is already one quarter of kcal on 36 atom system. > > 6. This isn't a conversion constant problem: conversion factor in G90 > is 1 Hartree = 627.5095 kcal/mole. (tokcal=627.5095) If we calculate > what this constant would be in MOPAC to give the same result as G90 > than tokcal = 630.7040 for MNDO and 630.1250 for AM1 method in the > same program, which is very unlikely. > > 7. Some people reported the same as the above MOPAC 6.0 > numbers on the following hardware and/or software: > HP/750 MOPAC 6.0 > SG/4Dgt70 AMPAC 3.01 > Unknown AMPAC 4.0, AMPAC 2.1 > > 8. I changed also position of first atom in benzene for 0.001 A and the > difference in energy was only 0.00196 kcal/mole. The difference in > porgrams is 0.01082. I consider 0.001 A significant change in geometry > in terms of roundoff errors. > > Conclusion: > > It seems to me that this isn't roundoff error nor conversion > factor problem, but either of the two: > 1. parameters for C and/or H atoms are different > 2. code is not implemented correctly > > Regards -- Milan Hodoscek > Regarding Milan Hodoscek's query about MOPAC: Have you tried rotating the molecule and seeing if the answers are invariant to rotations? This may have nothing to do with the current problem. However, I have heard from a few individuals (using older versions of MOPAC?) that rotational invariance was a problem. Just a thought. Mark Thompson From rhn900@anusf.anu.edu.au Tue Sep 22 08:24:16 1992 From: Ross Nobes Subject: Discrepancy in energies between MOPAC and Gaussian To: chemistry@ccl.net Date: Tue, 22 Sep 92 13:24:16 EST Milan Hodoscek has raised some questions regarding discrepancies in energies between MOPAC and Gaussian. This is caused by the different factors used in the two programs to convert between angstroms and bohrs - MOPAC 6 uses 0.529167 whereas the Gaussian implementation uses 0.52917706. Using MOPAC 6.01 and Gaussian 92 Revision C on a Fujitsu VP2200 I get the following results for a PRECISE, MNDO calculation on benzene: MOPAC 6 21.87987 kcal/mol Gaussian 92 (original code) 21.76893 Gaussian 92 (using MOPAC conversion factor) 21.87975 -- Ross Nobes R.Nobes@anu.edu.au Australian National University Supercomputer Facility GPO Box 4, Canberra ACT 2601, Australia Fax +61 6 2473425 Phone +61 6 2494154 / 4161