From BORKENT@caos.caos.kun.nl Mon Oct 4 10:08:35 1993 Date: Mon, 4 Oct 1993 10:52:08 METDST From: BORKENT@caos.caos.kun.nl Subject: entropy of methyl group To: CHEMISTRY@ccl.net Message-Id: <01H3PN4OEON694DS47@caos.caos.kun.nl> Dear netters, Is there somebody outthere who can educate me on the following: I'm using AMPAC to reproduce a racemization path, starting from the transition state, down to one enantiomer. In the molecule there is a methyl group, rotating (almost) freely in the TS and enantiomer, but hindered halfway. As an organic chemist I've always been told that this causes an entropy effect, increasing the energy of the system, maybe even creating another TS? So I would like to make this correction to the enthalpy (heat of formation) values obtained from AMPAC, but the question is, how? The manual states that a FORCE/THERMO calculation only makes sense in stationary points, not halfway. I can trace some frequencies of the CH3 oscillations, but can these be used to calculate an entropy contribution at any point of the curve? So far the problem, no need to elaborate further and to bore the general audience. I would be grateful to people giving this case a thought and I am of course willing to provide more details if needed. Please contact me directly to discuss the issue and I'll summarize anything useful. Hens Borkent CAOS/CAMM Center University of Nijmegen The Netherlands Tel +31-80-652137 borkent@caos.caos.kun.nl From chp1aa@surrey.ac.uk Mon Oct 4 10:36:06 1993 From: Mr Andrew D Allen Message-Id: <9310040957.AA28977@central.surrey.ac.uk> Subject: DFT: SUMMARY To: chemistry@ccl.net Date: Mon, 4 Oct 93 10:57:52 BST Here it is at last. Thanx to all who contributed, I will leave the summary on our local ftp server (131.227.110.2 : login anonymous, etc) for a while, in postscript and ascii form. Density functiional theory An introduction to DFT may be found in: J.Weber, H.Huber and H.P.Weber, Chimia vol 47, 57-59 (1993) Computational Chemistry: All you ever wanted to know about density functional theory. Dear Andy, Springer-Verlag published a couple of book on density functional methods. I attach the respective information for you. The book edited by J.K. Labanowski is in high demand and might be out of print (only few copies were available when I last looked into the sales figures). There have also been several contributions to Theor. Chim. Acta. Sincerely yours, Dr. Rainer Stumpe Chemistry Editorial Springer-Verlag Tiergartenstr. 17 D-69121 Heidelberg Phone: +49-(0)6221-48 73 10 Fax: +49-(0)6221-41 39 82 INTERNET:stumpe@spint.compuserve.com Lecture Notes in Chemistry Vol. 50 D. MUKHERJEE, Calcutta, India (Ed.) Aspects of Many-Body Effects in Molecules and Extended Systems Proceedings of the Workshop-Cum-Symposium Held in Calcutta, February 1-10, 1988 1989. VIII, 565 pp. Soft cover DM 128,-. ISBN 3-540- 50765-5 Contents: Separability in Many-Electron Problem; Size Extensivity and Size-Consistency. Many-Body Perturbation Theory and Coupled-Cluster Theory of Atomic and Molecular Electronic Structure. Propagators in Bound States and Resonances. Condensed Matter and Related Topics. Many- Body Methods in Dynamical Processes. Relativistic Methods with Applications. Group-Theoretic Techniques. Few-Body Methods, Large-N Expansion and Other Mathematical Topics. Densities and Density-Functionals. List of Participants. Subject Index. This volume features invited lectures presented in the workshop-cum-symposium on aspects of many-body effects in molecules and extended systems, Calcutta, February 1 - 10, 1988. The organizers invited leading experts to present recent developments of many-body methods as applied to molecules and condensed systems. The panorama portrayed is quite broad, but by no means exhaustive. The emphasis is undoubtedly on a "molecular point of view". U. Harms, Industrieanlagen-Betriebsgesellschaft (IABG) Ottobrunn, FRG (Ed.) Supercomputer and Chemistry IABG Workshop 1989 1990. Approx. 150 pp. 46 figs. 38 tabs. Softcover DM 98,- ISBN 3-540-52915-2 This volume represents the contributions of the 1989 IABG workshop on supercomputers and chemistry. >From the CONTENTS : Transputer Graphic Supercomputing MIMD Type Computers Drug Design Peptide and Protein Engineering Density Functional Calculations Computational Chemistry, Theoretical Chemistry, Drug Design, Protein Engineering, Super and Parallel Computing, Transputer Applications For industrial and academic researchers, research managers in industry and computer manufacturers and vendors in the above-mentioned fields J.K. Labanowski, Ohio Supercomputer Center, Columbus, OH; J.W. Andzelm, Cray Research Inc., Eagan, MN (Eds.) Density Functional Methods in Chemistry 1991. XII, 443 pp. 68 figs. Hardcover DM 128,- ISBN 3-540-97512-8 Predicting molecular structure and energy and explaining the nature of bonding are central goals in quantum chemistry. In this volume, which is based on a 1990 workshop at the Ohio Supercomputer Center, leading experts on the density functional (DF) method and its chemical applications demonstrate how their work contributes to these goals and has come into its own as an advanced method of computational chemistry. Density functional theory emerged as an alternative to the traditional ab initio and semiempirical approaches of quantum chemistry for studying the ground state properties of molecular systems. Advantages of the method are its high accuracy and efficiency, which make it particularly well suited for the realistic study of large molecular systems of practical importance. It describes with consistent reliability organic, in organic, metallic, and semiconductor systems consisting of elements throughout the periodic table. For these reasons, density functional methodology is used increasingly in pharmaceutical, agrochemical, and biotechnology research; materials and polymer science; catalysis, surface and solid state research; and electrochemistry and microelectro nics. The wealth of applications presented in this book will make it a valuable tool for researchers faced with a growing choice of software and theoretical approaches. Structure and Bonding Volume 80 Chemical Hardness K. Sen, University of Hyderabad, India; D.M.P. Mingos, Imperial College of Science, Technology and Medicine, London, UK (Eds.) With contributions by numerous experts 1993. Approx. 280 pp. 52 figs. Hardcover DM 228,- ISBN 3-540-56091-2 Contents: R.G. Pearson, Santa Barbara, CA: Chemical Hardness - An Historical Int roduction. P.K. Chattaraj, Karagpur, India; R.G. Parr, Chapel Hill, NC: Density Functional Theory of Chemical Hardness. J.L. Gazqu z, Mexico, Mexico: Hardness and Softness in Density Functional Theory.- L. Komorowski, Wroclaw, Poland: Har dness Indices for Free and Bonded Atoms. N.H. March, Oxford, UK: The GroundStat e Energy of Atomic and Molecular Ions and Its Variation with the Number of Elect rons. K. Sen, Hyderabad, India: Isoelectronic Changes in Energy, Electronegativ ity, and Hardness in Atoms via the Calculations of . P. Politzer, J.S. Mur ray, M.E. Grice, New Orleans, LA: Charge Capacities and Shell Structures of Atoms. R. F. Nalewajski, Cracow, Poland: Hardness Based Molecular Charge Sensitivit ies and Their Use in the Theory of Chemical Reactivity. B.G. Baekelandt, R. A. Schoonheydt, W.J. Mortier, Leuven-Heverlee, Belgium: The EEM Approach to Chemica l Hardness in Molecules and Solids: Fundamentals and Applications. J.A. Alonso, L.C. Balbas, Valladolid, Spain: Hardness of Metallic Clusters. Z. Zhour, P.K. Chattaraj, R.G. Parr, C. Lee First-order gradient correction for the exchange-energy density functional for atoms Theoretica Chimica Acta Volume 84 Number 3, p 237 P. Politzer, J. M. Seminario, M. C. Concha, J. S. Murray Some applications of local density functional theory to the calculation of reaction energetics Theoretica Chimica Acta Volume 85 Number 1-3, p 127 D. Heinemann, A. Rosen Basis-independent potential energy curves for the neutral diatomics of Li, Na and K evaluated by means of Hartree-Fock and different density functional potentials Theoretica Chimica Acta Volume 85 Number 4, p 249 Here are a few references which contain systematic comparisons of various properties of small molecules by DFT and conventional ab initio (HF, MP2, QCI) as well as experiment. Some of the references therein are also useful along these lines. I hope you find them helpful. B. G. Johnson, P. M. W. Gill and J. A. Pople, J. Chem. Phys. 98, 5612 (1993). A. D. Becke, J. Chem. Phys. 98, 5648 (1993). A. D. Becke, J. Chem. Phys. 97, 9173 (1992). A. D. Becke, J. Chem. Phys. 96, 2155 (1992). J. Andzelm and E. Wimmer, J. Chem. Phys. 96, 1280 (1992). Benny Johnson Department of Chemistry Carnegie Mellon University Message 39/53 from rec@ncifcrf.gov Aug 5 '93 at 12:04 (noon) Reply-To: cachau@ncifcrf.gov A good reference book is: "Density Functional Methods in Chemistry" Jan K. Labanowski & Jan W. Andzelm Ed. Springer-Verlag Berlin ISBN 3-540-97512-8 We have made a systemmatic comparison of polarizabilities and hyperpolarizabilities calculated at the HF level (with HONDO8), with 2 different DFT programs (deMon and DMol), and with experiment. The reference is ... J. Guan, P. Duffy, J.T. Carter, D.P. Chong, K.C. Casida, M.E. Casida, and M. Wrinn, J. Chem. Phys. 98 (1993) 4753. ``Comparison of local-density and Hartree-Fock calculations of molecular polarizabilities and hyperpolarizabilities'' The results are quite encouraging. An interesting point is that comparisons between HF and DFT with overly small basis sets (valence triple zeta in our case) are dominated by basis set effects instead of the treatment of exchange and correlation. Thus the quantitative advantages of DFT only became apparent as the basis sets became more complete. Of course, convergence rates will depend on the property. ... Mark E. Casida Return-Path: From CHUCK@psipsy.uct.ac.za Mon Oct 4 12:08:36 1993 Message-Id: To: "Wojciech Galazka" , chemistry@ccl.net From: "Marais, Charles F. , Dr" Date: Mon, 4 Oct 1993 09:27:49 SAST-2 Subject: Re: C.I. on MOPAC 6.0 problem > When using keyword 'C.I =n' (n - any valid number). I get results >with 'charge on system = 1', although charge of the molecule >calculated is 0 ! Yes, I saw this in 1991, and asked James Stewart about it - he said that there seems "to be a bug in MOPAC regarding C.I. = 4" , and said he'll look into it. I assume that became part of the changes in MOPAC 7/93, but I haven't installed any of those. It would be most helpful if someone could try it out with the new version.... 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 JKONG@ac.dal.ca Mon Oct 4 06:19:39 1993 Date: Mon, 04 Oct 1993 09:19:39 -0300 From: JKONG@ac.dal.ca Subject: Re: "POKING" OUTPUT FILES ON VAX/VMS To: chemistry@ccl.net Message-Id: <01H3PK03J27C000R2G@AC.DAL.CA> From: DAL1::JKONG 4-OCT-1993 09:18:22.46 To: IN%"fredvc@esvax.dnet.dupont.com" CC: JKONG Subj: RE: "POKING" OUTPUT FILES ON VAX/VMS You can try to open the output file as SHARED. This should allow more than one processes to open the file at the same time. Good luck! Jing From jabs@chemie.uni-halle.d400.de Mon Oct 4 14:44:20 1993 Date: Mon, 4 Oct 1993 13:44:20 +0100 From: jabs@chemie.uni-halle.d400.de Message-Id: <931004134419*/S=jabs/OU=chemie/PRMD=UNI-HALLE/ADMD=D400/C=DE/@MHS> To: chemistry@ccl.net Subject: normal coord. Hi, i am looking for normal coordinate analysis for small alcohols (methanol, ethanol, 2-propanol, 2-phenylethanol) with ab initio or semiempiric (PM3/AM1) methods. Can anybody help me with literature or some results ? Thanks in advance Andreas From mckelvey@Kodak.COM Mon Oct 4 04:50:35 1993 Date: Mon, 4 Oct 93 08:50:35 -0400 Message-Id: <9310041250.AA19118@Kodak.COM> From: mckelvey@Kodak.COM To: osc@Kodak.COM Subject: scanning output from unit 6 on VMS... If the job is submitted to batch, then do not explicitly open unit 6. As a result, unit-6 info will appear in the logfile bearing the name of the batch-com. That file can be searched readily, or even just copied, I believe. On a unixmachine a similar result wil occur...if the nohup command is used, then unit-6 info will appear in nohup.out. John McKelvey From mikes@bioch.ox.ac.uk Mon Oct 4 15:10:53 1993 Date: Mon, 4 Oct 93 14:10:53 +0100 From: mikes@bioch.ox.ac.uk Message-Id: <9310041310.AA07959@nmrpcd.ocms.bioch.ox.ac.uk> To: chemistry@ccl.net Subject: Cluster Analysis Dear Comp-Chemers, Does anybody know how one goes about dividing 100 structures (say) with the same amino acid sequences into groups of similar structures. The way I would envisage doing this would be to say that two structures are in the same group if their rmsd (say) is less than 1 Angstrom, and that this relationship is transitive, but this is only guessing! Mike Smith From gl@beta.mdy.univie.ac.at Mon Oct 4 14:26:01 1993 Message-Id: <199310041358.AA00278@oscsunb.ccl.net> From: Gerald Loeffler To: CHEMISTRY@ccl.net Subject: SUMMARY: force fields for biomolecules Date: Mon, 4 Oct 93 14:13:44 MEZ On September 28 I asked what force fields people were using for the Molecular Dynamics study of biomolecules. (original posting appended to this mail) I like to thank David Case Arne Elofsson Peter Reinert and Don Williams very much for their replies! Here is what they conributed: ============================== David Case: (case@scripps.edu) ============================== uses AMBER and AMBER/OPLS for proteins. Implemented in Tripos, Discover, Chem-X, AMBER, SPASMS, WESDYN codes. An improved version is due to come out by the end of 1993. He also points to the newer all-atom CHARMM force fields, e.g. CHARMM22 implemented in CHARMM only. Older CHARMM force fields are used in X-PLOR, moil, md92. He considers it very hard to compare force fields but points to an article in the November issue of Chemical Reviews by Charles Brooks and David Case. ======================================== Arne Elofsson: (arne@ewald.mbi.ucla.edu) ======================================== thinks that other parameters than the force field are more important, like correct treatment of electrostatic cutoffs, switching/shifting and water implementations. He cites two papers by Guenot & Kollman: Protein Science 1992, 1185 - ? and Journal of Comp. Chem. 14, 3, 295 - 311, 1993 and three papers by Schreiber/Steinhauser -- which are in fact from our group that's why I don't cite them here. ================================================== Peter Reinert: reinert@vax.mpiz-koeln,mpg.d400.de) ================================================== thinks that AMBER is the most popular force field for biomolecules, the older version of which (AMBER 3.x) is implemented in HYPERCHEM, SYBYL, PROSIMULATE, INSIGHT/DISCOVER) The reference is: Weiner, S.J. et al "A new force field ... " JACS, 106, 765 - 784, 1984 He also expects GROMOS to be upgraded. ============================================ Don Williams: (williams%xray2@ulkyvx.bitnet) ============================================ points to a commercial program that calculates net atom charges and other electric multipoles from the molecular electric potential obtained by ab initio calculations. ===================================================================== If you would like to add some comments, here is the original posting: ===================================================================== Dear Chemist! We have been using the GROMOS force field implemented in the program gromos itself and in a home-made program for the MD-simulation of proteins in water for a long time. Since I consider this force field somewhat dated by now, I am asking you for your experience with FORCE FIELDS FOR BIOMOLECULES. I would very much appreciate answers to the following questions: 1) what is the force field you would recommend and which program you are using implements this force field 2) why would you recommend this force field - e.g. are there published comparisons between different force fields? 3) is source code for a program using this force field available? What I ideally would like to get is a reference to a paper describing the force field in detail and a 'sample implementation' by the developers of the force field including source code. Using that I will probably write a program that uses this force field on my own, since we are usually doing a lot of algorithmic tuning. I THANK YOU VERY MUCH FOR YOUR RESPONSES IN ADVANCE, Gerald -- +------------------------+ +-----------------------------------+ |Gerald Loeffler | |Theoretical Biochemistry Group | |gl@beta.mdy.univie.ac.at| |Department of Theoretical Chemistry| +------------------------+ |University of Vienna | +-----------------------------------+ +--------------------------------------------------------------------------+ |Institut fuer Theoretische Chemie und Strahlenchemie der Universitaet Wien| |Arbeitsschwerpunkt Theoretische Biochemie | |Waehringerstrasse 17/Erdgeschoss | |A-1090 Wien, Austria | +--------------------------------------------------------------------------+ From wgalazka@chem.uw.edu.pl Mon Oct 4 09:57:58 1993 Organization: Department of Chemistry, University of Warsaw From: "Wojciech Galazka" Date: Mon, 4 Oct 93 15:57:58 CST Message-Id: <621.wgalazka@zoolook.chem.uw.edu.pl_POPMail/PC_3.2.3_Beta_2> To: chemistry@ccl.net Subject: Semiempirical programs with d,f orbitals Hi Does anybody know if there are any quantum semiempirical programs with d, f orbitals included? The only one I know is SINDO1 (for reference see K.Jug R.Iffert J.Schulz 'Development and Parametrization of SINDO1 for Second-row Elements' Int.J. Quant. Chem. 32 265-277 (1987)}. I have heard about MNDO-like methods with artificially included d orbitals but results obtained by them are not good for hypervalent sulphur compounds. //////////////////////////////////////////////////////////// // // // Wojciech Galazka // // Computer Center // // Chemistry Department, University of Warsaw // // Pasteura 1, 02-093 Warsaw, Poland // // // // wgalazka@zoolook.chem.uw.edu.pl // //////////////////////////////////////////////////////////// From combariz@rouge.phys.lsu.edu Mon Oct 4 05:55:58 1993 Date: Mon, 4 Oct 93 10:55:58 CDT From: combariz@rouge.phys.lsu.edu (Jaime Combariza) Message-Id: <9310041555.AA16669@rouge.phys.lsu.edu> To: CHEMISTRY@ccl.net Subject: SCRF for ions Hi all: I have been working with Gaussian 92 and Gamess using the SCRF method to study solvated anions. First, I found out that the two programs would give me different results for similar input decks!!! A closer look revealed that this difference is roughly equal to the molecular Born energy, which apparently is not accounted for in the Gaussian implementation of the self-consistent-reaction-field. Question for all of you G92 users/gurus: Is this right? If it is, especifically for ions, should one not account for this energy? I should add that in some small test cases this energy is about 40 Kcal/mol. I will appreciate your comments, Jaime E. Combariza From leoh%lemoyne.BITNET@phem3.acs.ohio-state.edu Mon Oct 4 10:44:04 1993 Date: Mon, 04 Oct 1993 14:44:04 -0400 (EDT) From: "Leo, Howard" Subject: MOPAC6 FOR PC? To: CHEMISTRY@ccl.net Message-Id: <0097384D.5F4F9F80.31956@lemoyne> Dear Netters, Does anyone have, or know of, a ported version of MOPAC 6 for a PC? Howard Leo LEOH@LEMOYNE From EDGECOMK@QUCDN.QUEENSU.CA Mon Oct 4 11:22:00 1993 Message-Id: <199310041924.AA05870@oscsunb.ccl.net> Date: Mon, 4 Oct 1993 15:22 EDT From: EDGECOMK@QUCDN.QueensU.CA To: chemistry@ccl.net Subject: DFT-HF comments Jing is right about the boundary conditions with DFT, however, I think people are trying to work around that in various ways... One problem with HF is of course basis set selection and limits... then to get correlation MP theory also is nonvariational (illustrated quite well by Nick Handy at last years NATO ASI in Vancouver). The pros and cons can be bothersome. For example, DFT does not give you a 'wavefunction' you can put to one side and look at later... you have to calculate everything then and there. Of course, some will argue that as DFT is much faster so you just recalculate everything again... Also, what about excited states?... a sore point with DFT. WHich code and approximations are best?... are results found with one approximation (functional) comparable to those found with another? Once again the problem is choosing the method, code, approximations etc that suits the research problem. DFT may be able to give results in some areas that HF just cannot handle. However, coming back to some discussion that took place earlier, maybe for some systems of interest (polypeptides etc) MM is best! Just some thoughts... Ken Edgecombe Dept. of Chemistry Queen's University Kingston, ON CDN From adit@Kodak.COM Mon Oct 4 11:51:42 1993 Date: Mon, 4 Oct 93 15:51:42 EDT From: adit@Kodak.COM (Adi Treasurywala) Message-Id: <9310041951.AA02712@bcc9.kodak.com> To: chemistry@ccl.net Subject: MM/MD WORKSHOP ANNOUNCEMENT. FIRST ANNOUNCEMENT FOURTH BIENNIAL WORKSHOP ON MOLECULAR MECHANICS AND MOLECULAR DYNAMICS Sponsored by The Supercomputer Computations Research Institute Florida State University Tallahassee Florida Monday Feb 21 to Friday Feb 25 1994 Plenary Speaker who have accepted invitations so far: D.L.Beveridge Jack Dunitz Thomas Halgran Warren Hehre Kendall Houk Adi Treasurywala Invited Speakers who have accepted invitations so far: J.E.Anderson, J.Phillip Bpwen, Helena Dodzuik, Frank Leusen, Wayne Mattice, Dora Schnur, Terry Stouch, S.Swaminathan, Henk van der Plas, Bastian van de Graaf, David M. Gagne, David C. Doherty, Douglas A. Smith, Hubert Bodot. TOPICS: New developments in the quantitative computational modeling of molecular structure, molecular dynamics, energies (thermodynamics) including solvent effects, polymers, including biopolymers, inorganic systems. It is expected that the studies will reflect new approaches in the use of molecular mechanics or molecular dynamics rather than applications of well-developed techniques. Developments in the use of ab initio calculations in modeling are appropriate. PROGRAM: Monday evening: Reception. Tuesday through Friday noon: Scientific sessions Tuesday through Thursday: Vendor exhibits combined with electronic posters (see below) POSTERS: Contributed conventional posters will be displayed. ELECTRONIC POSTER SESSION:Please send all requests to take part in this new event no later than October 31 1993 to Adi M Treasurywala,Sterling Winthrop Inc,1250 South Collegeville Road, PO Box 5000, Collegeville, PA 19426-0900,Voice (215)983-6610 FAX (215)983-5559, INTERNET adit@kodak.com stating the software, hardware and approximate time requirements. ALL OTHER INQUIRES TO DELOS DETAR, DEPT. OF CHEMISTRY, FLORIDA STATE UNIVERSITY, TALLAHASSEE FL 32306-3006 TEL (904)644-3709. FAX (904)644-8281 detar@mailer.fsu.edu Acceptance cannot be guaranteed. Deadline January15 1994. CONTRIBUTED TALKS: A few additional contributed talks may be accomodated. FEES: The workshop fee is $350 ($325 if received before January 15 1994). For those whose talks/posters have been accepted the fees will be $250. Student rate $150. Fees will cover a hospitality reception on March 26 continental breakfasts 4 days lunches 3 days coffee breaks and conference banquet. Registration Form Molecular Mechanics and Molecular Dynamics 1994 Program # 1902994 To register, please complete this form and mail it with your payment (checks made payable to Florida State University) to: Conference Registrar, Florida State Conference Center, Florida State University, Tallahassee FL 32306-2027. For registration information, please call 904-644-3806. For information about the program, please call Pat Meredith at 904-644-1866, or by e-mail at meredith@scri.fsu.edu.. Name:____________________________________ Social security #: _______________________(This is not essential, but will speed the process if you require a reimbursement of fees) Organization: _________________________________________ Address: _____________________________________________ _________________________________________________ Phone: ______________________ Fees: $325 before January 15, 1994 $350 after January 15, 1994 $150 for students $250 for those whose contributed talks or posters are accepted I wish to pay my fee of $______by ___MasterCard, or ___Visa Acct. # _________________________ Exp. Date ________________ Signature: __________________________________________ Please note, the university adds a 2% service fee for credit card payments. From st-amant@theory.chem.uottawa.ca Mon Oct 4 14:52:29 1993 Date: Mon, 4 Oct 1993 18:52:29 -0400 From: st-amant@theory.chem.uottawa.ca (alain st-amant) Message-Id: <9310042252.AA04953@theory.chem.uottawa.ca> To: chemistry@ccl.net, mikes@bioch.ox.ac.uk Subject: Re: Cluster Analysis > Dear Comp-Chemers, > > Does anybody know how one goes about dividing 100 structures (say) with the > same amino acid sequences into groups of similar structures. > > The way I would envisage doing this would be to say that two structures are > in the same group if their rmsd (say) is less than 1 Angstrom, and that this > relationship is transitive, but this is only guessing! > > Mike Smith A very nice piece of work has recently come out of Charlie Brooks' group. It should be just what you want. The reference is: M. E. Karpen, D. J. Tobias, and C. L. Brooks, "Statistical Clustering Techniques for the Analysis of Long Molecular Dynamics Trajectories -- Analysis of 2.2 ns Trajectories of YPGDV," Biochemistry, Volume 32, pages 412-420. Sincerely, Alain St-Amant st-amant@theory.chem.uottawa.ca From shenkin@still3.chem.columbia.edu Mon Oct 4 15:04:10 1993 Date: Mon, 4 Oct 93 19:04:10 -0400 From: shenkin@still3.chem.columbia.edu (Peter Shenkin) Message-Id: <9310042304.AA06034@still3.chem.columbia.edu> To: mikes@bioch.ox.ac.uk, chemistry@ccl.net Subject: Re: Cluster Analysis > From: mikes@bioch.ox.ac.uk > Does anybody know how one goes about dividing 100 structures (say) > with the same > amino acid sequences into groups of similar structures. > The way I would envisage doing this would be to say that two structures > are in > the same group if their rmsd (say) is less than 1 Angstrom, and that this > relationship is transitive, but this is only guessing! I'm glad you asked that question. :-) Quentin McDonald and I have written a program that does exactly what you describe: cluster analysis of molecular conformations. An article that covers the approach we used, as well as the implementation, has been submitted to J. Comput. Chem. The program is called XCluster. It begins by constructing the matrix of inter-conformational "distances" between all pairs of conformations read in. There are several choices of "distance" available, including RMS of interatomic distances following rigid-body superposition (which is what I assume you mean by "rmsd"). Molecular symmetry can be taken into account, and clustering can, if desired, be performed on only a subset of the atoms -- for example, the ring atoms of a cyclic system. > The way I would envisage doing this would be to say that two structures > are in > the same group if their rmsd (say) is less than 1 Angstrom, and that this > relationship is transitive, but this is only guessing! The relationship does satisfy the triangle inequality, but I don't think the term "transitivity" is applicable, if I remember its definition correctly (which I may not :-) ). A very nice proof of the satisfaction of the triangle inequality for RMS interatomic distance following best rigid-body interatomic superpostion is given in our paper. The proof is due to Dr. Mathis Thoma of the CIBA-Geigy Corporation. XCluster allows you to perform clustering based either on an a-priori distance criterion (eg, the 1 Angstrom RMS distance you propose), or else based on statistical criteria (some of which we devised ourselves) for determining whether there is some "natural" clustering distance inherent in the data. Our statistical criteria are imperfect, but they work in many cases -- especially where the clustering is especially clear-cut. XCluster has been distributed along with the molecular modeling package MacroModel since Release 4.0 of MacroModel, which came out in July. The package does cost money, but is inexpensive to academic institutions. For pricing, availability and an overview of its capabilities, issue the following command:: finger mmod@still3.chem.columbia.edu Sorry, but XCluster is available only with MacroModel. As the name implies, XCluster has a spiffy X interface, including some nice visualization tools (eg, visualization of the inter- conformational distance matrix). XCluster can also write out the clusters of conformations, superimposed and colored by cluster, in MacroModel format. This format is understood by several other programs as well as MacroModel itself. XCluster is a standalone program, but it also can "talk to" MacroModel. For example, if you display a compound in MacroModel, the set of atoms to be used in the cluster analysis may be "picked" with the mouse, and thereby transmitted to a simultaneously running XCluster program. XCluster can also run in a mode which accepts not molecular structures, but rather the N(N-1)/2 non-redundant distance-matrix elements, then applies the cluster-analysis to that matrix. Thus it can be used to visualize and analyze arbitrary distance data. We have additional facilities planned for future releases. I hope I have not been "tooting my own horn" too loudly in this posting. Obviously, I have an intimate association with MacroModel and with XCluster, and am thrilled to hear of interest in the subject of clustering of molecular conformations. -P. ************************f*u*cn*rd*ths*u*cn*gt*a*gd*jb************************ Peter S. Shenkin, Box 768 Havemeyer Hall, Dept. of Chemistry, Columbia Univ., New York, NY 10027; shenkin@still3.chem.columbia.edu; (212) 854-5143 ********************** Wagner, Beame, Screvane in '93! ********************** From states@ibc.wustl.edu Mon Oct 4 13:26:37 1993 Date: Mon, 4 Oct 93 18:26:37 CDT From: states@ibc.wustl.edu (David J. States) Message-Id: <9310042326.AA10458@ibc.WUStL.EDU> To: chemistry@ccl.net, mikes@bioch.ox.ac.uk Subject: Re: Cluster Analysis |> Does anybody know how one goes about dividing 100 structures (say) with the same |> amino acid sequences into groups of similar structures. |> |> The way I would envisage doing this would be to say that two structures are in |> the same group if their rmsd (say) is less than 1 Angstrom, and that this |> relationship is transitive, but this is only guessing! |> |> Mike Smith There are several issues here. If you do not know how many clusters there should be, then you don't want to use an algorithm that imposes a particular answer. This can happen either explicitly (for example a binary classification halted at 3 divisions by definition will give 8 classes) or implicitly in a leader-mean classification of the sort you describe (the number of classes is inversely dependent on the cutoff radius). To avoid biasing the results of the classification by class number, you need a method that allows you to compare classifications with different numbers of classes. This leads to the general field of Bayesian classification where a classification is viewed as a model for the observed distribution and the optimal classification is that classification which optimally describes the data. Larry Hunter and I used this to derive a classification of protein secondary structure several years ago (Hunter and States (1991), "Bayesian classificaiotn of protein structural motifs.", in Proceedings of HICSS-24, IEEE Press, Los Alamitos CA, 595-604). Another issue is flat vs. hierarchical classification. Are these proteins derived by an evolutionary process from a common ancestor in which case a hierarchical model might be more appropriate, or are they random samples of conformation space in which case a flat classification would be better. Defining the optimal tree structure for a set can itself be a demanding problem. The computational complexity of the problem depends on your class definition. If you seek connected classes (two structures are in the same class if there is a path of similarity relationships connecting them, transitive closure) then this is algorithmically a minimal spanning tree problem for which linear time solutions exist. On the other hand, if you demand that all members of a class to fall within some cutoff of every other member (cliques), then you have a graph partition problem that is NP-complete. The sensitivity of the classification to errors or ambiguities in the data is also dependent on class definition. Transitive closure is robust to less than perfect sensitivity in defining similarity relationships but false positive similarity judgements lead directly to classification errors. On the otherhand, clique definitions are very sensitive to false negative similarity judgements and robust to false positives. David States Institute for Biomedical Computing / Washington University in St. Louis From mercie@med.cornell.edu Mon Oct 4 15:37:18 1993 Date: Mon, 4 Oct 1993 19:37:18 -0400 (EDT) From: Gustavo Mercier Subject: DFT To: chemistry@ccl.net Message-Id: Hi, Netters! Recently there have been a few questions about DFT. Although I don't consider myself an expert, I have been listening to people talk about DFT for a few years and recently started to do DFT computations on metalloporphyrins. I hope the following will clarify some points. I certainly welcome any corrections of my statements below! DFT (Density Functional Theory) stands as a reformulation of the Schroedinger Equation. Its origin really dates from the early days of quantum mechanics (Thomas-Fermi-Dirac, Slater's work, etc), but its "modern" foundation is based on the theorems of Hohenberg and Kohn developed in the '60's, and its practical implementation in Quantum Chemistry rests on the Kohn - Sham equations. For a beautiful and concise description of the historical development of DFT I suggest you read the article by Hohenberg, Kohn, and Sham in Advances of Quantum Chemistry v. 21 special ed. S. Trickey pp 7 -26, 1990. The HK theorems essentially state: 1) Given a density, rho, the external potential, Vext, is fixed. For Quantum Chemistry using the electronic molecular hamiltonian within the Born-Oppenheimer approximation, the Vext is the electrostatic potential generated by the nuclei. 2) The energy is a UNIQUE Functional of the density. As originally described, the above applies only to NON-DEGENERATE GROUND STATE systems! Following their work, issues described as the V-representability and N-representability problems were identified and dealt with in a variety of ways. For an explanation of these and the rigourous foundation of DFT theory, try Kryachko and Ludena, Energy Density Functional Theory of Many-Electron Systems by Kluwer Academic Publishers Understanding Chemical Reactivity Series v. 4, 1990. For computational chemists the Kohn - Sham equations are the origin of most implementations of DFT theory. Essentially they apply a variation to E[rho] = F[rho] + Eext[rho]; F = Ts + Eee + Exc where Eee is the coulombic electron - electron repulsion and Exc is the "exchange - correlation" term, a functional of rho. The key point to understand is Ts. Ts is the kinetic energy for a collection of NON-INTERACTING electrons!. As you can see the structure is similar to our familiar Hartree - Fock, but there are subtle differences. In Hartree - Fock, the Kinetic term is for an INTERACTING set of electrons. The difference is THROWN into the undefined Exc term. KS showed that if Exc is known exactly, the KS equations will yield the EXACT density! The KS equations are generated using the above functional and the HK variational principle that stems from Theorem #2. The key result is that due to the choice of reference state, the KS equation is identical to the HARTREE equation, the Schroedinger Eq. for a system of non-interacting electrons, but with a new potential: Veff = Vext + Vee + Vxc Following the Hartree Eq., the density can be written EXACTLY as rho = Sumi phi(i)^2 This form of rho is NOT an approximation as is the case when the density is generated from the HARTREE - FOCK Equation. The rho looks identical in form, but the phi's are very different!! In solving the KS equations a Variational Method is used that includes Lagrangians due to the constraint of the density to integrate to the total N number of electrons. In Hartree - Fock, these correspond to the orbital energies as shown through Koopman's Theorem. In DFT, the Lagrangians do not have the same meaning. THEY DO NOT correspond to orbital energies. Through some manipulations you can generate "orbital energies" but they do not come explicitly from the KS equations. The fact that orbital energies are not generated also means that the phi's above don't have the same meaning they have within HF theory! This is one of the most distressing things for us chemist who always think in terms of orbitals!!! Only the density has "meaning", not its "components". The physical interpretation associated with the orbitals in HF theory which is useful in the generation of different configurations for CI or GVB computations is lost! Most of the work in DFT deals with defining Vxc. If it were known exactly, we would have the exact equation and an exact solution within the basis set expansion we chose for the phi's. In other words, within Hartree - Fock we have an exact operator with its approximate solution (single determinant), but in DFT we have an approximate functional with its exact solution. The ADVANTAGE is that the approximate part is small, and the density is a function of only three variables! In fact, using Mathematica and the output of my DFT runs, I have written the analytic expression for the the density of a 66 electron system using a double zeta + polarization basis set! All computed in an INDIGO R4000. It is satisfying to actually see the analytic form of your result! Given the functional associated with a molecular property, you can easily compute the property. Particularly, standard numerical algorithms can fairly easily be applied since the dimensionality of the problem has decreased significantly! For example, in my case from 66*3 to 3. When does DFT - KS fails? When your choice of Vxc or basis set is bad! Also, from the practical point of view, many programs expand the the electron density used to compute the Vee. If this basis is poor you also get bad results. One thing that has been appreciated is that if you want to adequately reproduce H-bonds, a critical point for the many biochemically oriented people in this mailing list, you need good descriptions of Vxc that include gradient expansions etc. Sorry for the bandwith, but I hope the above was useful. good luck mercie@cumc.cornell.edu From cramer@maroon.tc.umn.edu Mon Oct 4 16:21:34 1993 Message-Id: <0012cb0da2e011938@maroon.tc.umn.edu> From: "Christopher J Cramer-1" Subject: Re: SCRF for ions To: combariz@rouge.phys.lsu.edu (Jaime Combariza) Date: Mon, 4 Oct 93 21:21:34 CDT Jaime, > > Hi all: > > I have been working with Gaussian 92 and Gamess using the SCRF method > to study solvated anions. First, I found out that the two programs > would give me different results for similar input decks!!! > A closer look revealed that this > difference is roughly equal to the molecular Born energy, which > apparently is not accounted for in the Gaussian implementation of the > self-consistent-reaction-field. > Your analysis is exactly correct, the monopole (or Born) term is not included in the multipole expansion (which is truncated at the dipole in the Onsager model). In principle, one can take the G92 calculated spherical radius "a" that was used in the SCRF and use it to calculate the Born free energy from the classical Born eqn G = -1/2 ( 1 - 1/dielectric) q^2/a, where q is the charge on the molecule (all units atomic units). Of course, in a charged molecule, the dipole is no longer origin independent, and one needs to worry a bit about where the center of the sphere is being placed -- center of mass? center of charge? No matter what, the Born-Onsager model in a spherical cavity approximation starts to get pretty risky about this point. Of course, it's fast and simple. CJC -- Christopher J. Cramer University of Minnesota Department of Chemistry 207 Pleasant St. SE Minneapolis, MN 55455-0431 (612) 624-0859 cramer@maroon.tc.umn.edu From mail Mon Oct 4 16:50:39 1993 Date: Mon, 4 Oct 1993 16:19:02 -0400 From: hyper!ostlund (Neil S. Ostlund) Message-Id: <9310042019.AA19423@hyper.hyper.com> To: chemistry@ccl.net Subject: conjugation in mol mech John McKelvey has commented here about torsional angles in biphenyls. HyperChem also gives reasonable values of torsional angles in molecules such as this, but the issue is a very important one that is fundamental to the success or failure of molecular mechanics approaches and although I rarely contribute to discussions here, I thought that I might briefly comment on this topic. The standard molecular mechanics methods (MM2, Amber, CHARMm, etc) assign parameters on the basis of the "atom type" of the relevant atoms involved (4 atoms,in the case of a torsion) without consideration of the bond type. This results in the same torsional constants for the SINGLE BOND in biphenyl at for the AROMATIC RING BONDS of Benzene!! Thus, all the standard methods mentioned above result in biphenyl being planar which is unfortunate chemistry. The molecular mechanics secret is to recognize more of the chemical environment of a bond torsion than just the 4 "atom types". Alternatively, one might expand the number of different atom types, but this has its own problems. For example, in biphenyl it is important that one is trying to describe the torsion of a "single" bond (bond-order=1) not an "aromatic" benzene bond (bond-order=1.5). Chemical ideas like this are not recognized by the simple "atom type" methods, but are recognized by Dreiding, MM+, and others. In HyperChem, the MM+ method does standard MM2 calculations when explicit parameters are available in the parameter file associated with the relevant atom types. Thus a "standard" calculation results in planar biphenyl for the reasons described above. However, the MM+ force field in HyperChem falls back to a Dreiding-like scheme when explicit parameters for the torsion in question are not available in the parameter file. It then uses information about the "bond type" to derive default parameters. The result is that you get a better result when you don't have an explicit MM2 parameter and the default scheme is used! This points out one of the major defficiencies of standard molecular mechanics procedures like MM2, Amber, and CHARMm - they don't recognize the "bond type" of a torsion but only the "atom type" of the 4 relevant atoms. For those familiar with HyperChem, if you select the molecule and set all the atoms types to unknown (**) to dismiss the standard MM2 parameters and fall back to the default scheme, you will find reasonable geometries for systems like biphenyl, butadiene, etc. I don't want to sound as though I am promoting HyperChem here; I only use it as an example of an important issue for molecular mechanics calculations. Other programs also recognized this problem associated with the simple "atom type" approach. ------------ Neil Ostlund President, Hypercube Inc. 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 internet: ostlund@hyper.com