From d3g359@rahman.pnl.gov Thu Sep 30 13:27:40 1993 Date: Thu, 30 Sep 93 20:27:40 -0700 From: d3g359@rahman.pnl.gov Subject: Reply to Doug Smith To: chemistry@ccl.net Message-Id: <9310010327.AA06904@rahman.pnl.gov> In regard to Doug Smith's recent comment: Considering the recent huge increases in computation power and reduction in cost of workstations, all (hopefully) computational chemists will probably see some significant increase in the amount of cpu power at their disposal. Wouldn't it be great if we simply used this power to solve more and more real problems in chemistry, (an thus make ourselves more useful) rather than just trying to do bigger calculations? Aren't there a bunch of interesting problems that don't require huge basis sets and tons of correlation left to tackle? Just a thought. By the way, I'm not claiming that I'm not also guilty of this "bigger is better" mindset. ------------------------------------------------------------------------------ John Nicholas Senior Research Scientist Molecular Science Research Center Battelle Pacific Northwest Laboratory Mailstop K1-90 Richland, WA 99352 (509) 375-6559 ------------------------------------------------------------------------------ From chris@glycob.ox.ac.uk Fri Oct 1 05:19:03 1993 Date: Fri, 01 Oct 1993 09:19:03 EDT From: chris@glycob.ox.ac.uk To: chemistry@ccl.net Message-Id: <009735C4.7857EF40.9707@glycob.ox.ac.uk> Subject: Summary of simulated annealing replies Dear netters, Below is a summary of the replies I received with regard to my enquiry about simulated annealing. Tom Bishop and Mark Thompson wrote that the hot solvent/cold solute problem was due to cutoff problems in the system. Reference was made to the work by Guenot and Kollman (J. Comp. Chem. 14(3) 295-311 (1993)) on the problems of truncation of the nonbonded interactions. Tim Harvey wrote that the problem appeared to arise because of the shorter rotational correlation time of the solvent. A.P. Heiner made the point that with the Berendsen temperature coupling algorithm, a centre-of-mass motion is generated in the system over a long simulation time and that this must be subtracted from the motion of the system. Claudio Soares suggested using a coupling constant of 0.01 in the first ps of the simulation and 0.1 thereafter to obtain a slower rate of heating. Slower cooling rates were also suggested. Johan Postma gave this reply: As far as I remember, when I was in Groningen we used to have two different coupling constants to the thermal bath when doing MD on a protein in solution. A rule of thumb was to have a more tight coupling constant for water. (typically 0.05 - .1 for water and .2 - .4 for the protein). It all depends on the time step used and how fast one wants to do the cooling. The scaling factor is calculated as: s = 1 + t / c ( T0 / T - 1) where: t = timestep c = coupling constant T = temperature T0= bath temperature In order not to perturb the system too much s should not deviate too much from the value one.. (i.e. instead of cooling from 1000K to 300K one could also to this in steps 1000K -> 800K ; 800K - 600K or let the temperature of the surrounding bath drop with some timeconstant. Thank you to all these contributors and to others who replied to my question. I hope that I have not misrepresented anyone's views. *************************************************************** * Dr. Chris. Edge * Tel. +44-865-275-338 * * Glycobiology Institute * Fax. +44-865-275-216 * * Oxford University * e-mail chris@glycob.ox.ac.uk * * South Parks Road * 100117.3646@compuserve.com * * Oxford OX1 3QU U.K. * * *************************************************************** From cramer@maroon.tc.umn.edu Fri Oct 1 03:47:02 1993 Message-Id: <0012cac34d6016583@maroon.tc.umn.edu> From: "Christopher J Cramer-1" Subject: van der Waals radii To: chemistry@ccl.net Date: Fri, 1 Oct 93 8:47:02 CDT For thought: Gustavo A. Arteca recently posted some commentary which I have been thinking a lot about lately, and I'd like to offer my own paraphrase. van der Waals radii are not physical constants, they are parameters. I make this point because I see a lot of modeling done which involves taking the "van der Waals surface" and evaluating some property there. While this particular property may have nothing to do with the physical measurements which were used to develop the vdW radii from some OTHER physical model, nevertheless they are justified as somehow being "better" than parameters which are in fact optimized for EXACTLY the property being studied. My own personal irritation arises when people use Bondi's values, which are derived from X-ray analysis of crystal packing, to create solute cavities in continuum solvation modeling, and seem to regard this as the only God-given legitimate cavity which may be used. Actually optimizing better radii to fit solvation free energies is regarded as somehow "cheating". It makes me wonder what Bondi did to achieve such divine status . . . Perhaps to end on an amusing note, I have always found it fascinating that Gaussian's SCRF model develops a spherical cavity radius by calculating the 0.001 au isodensity surface, scaling the enclosed volume by 1.44, taking the radius which would deliver that volume for a sphere, and then adding 0.5 angstroms. Hey, if Gaussian is willing to be parametric, who am I to say nay? CJC P.S. While my own personal prejudice is showing, naturally I am interested in hearing other's opinions. P.P.S. The Gaussian work is primarily published by Wiberg, Wong, and Frisch, and is quite nice -- no slur is implied or intended. -- 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 jaeric@mtcamm.monsanto.com Fri Oct 1 06:44:45 1993 From: Jon A. Erickson Message-Id: <9310011644.AA14663@mtcamm.monsanto.com> Subject: charge transfer summary... To: chemistry@ccl.net (OSU Comp. Chem. List) Date: Fri, 1 Oct 93 11:44:45 CDT The following is a summary of the responses to my question about charge-transfer calculations. Thanks to all who responded and sorry about the delay in the posting of this summary. original posting: > Does anyone know of any references, programs, or other info > on how to deal with geometries and energies of charge-transfer > complexes? Currently, the approach I'm taking is semi-empirical- > CI methods. responses: ------------------------------------------------------------------------- From: Anthony Stone You need to beware of basis set superposition error, which vitiates many low-level calculations of charge-transfer interactions. Most so-called charge-transfer complexes are bound primarily by electrostatics, and the `charge-transfer interaction' makes a relatively small contribution. See e.g. Buckingham, A. D. & Fowler, P. W. J. Chem. Phys. (1983) 79, 6426-6428. Do electrostatic interactions predict structures of van der Waals molecules? Canad. J. Chem. (1985) 63, 2018-2025. A model for the geometries of van der Waals complexes. and Stone, A.J. Chem. Phys. Letters (1993) 211, 101-109 Computation of charge-transfer energies by perturbation theory ------------------------------------------------------------------------- From: yong@msi.ch (Dr. Yong Li Application Scientist) Jon Erickson asked for materials dealing with charge-transfer complexes: The "Charge Equilibration" methode developped by Rappe and Goddard (J.Phys. Chem.,95, 3358, 1991) can be used for such complexes. This methode is available in software packages CERIUS and POLYGRAF provided by Molecular Simulations Inc. Combined with the force field methode in these packages, you should be able to study the geometries and energies of your complexes. ------------------------------------------------------------------------- From: afj@chem.ucla.edu (Andy Jacobson) I asked a similar (more specific question to the list some time ago. I am enclosing here my summary. I hope this is of some use. Cheers, A.J. Dear Comp Chemers, A couple of weeks ago I requested info on MD work on TICT states in molecules containing aryl-NR2 groups. A large number expressed interest in the results, so I am posting the results to chem.ccl.net .....Unfortunately I received far fewer replies then inquiries, three to be exact. Two of these were on the subject..... ++++++++++++++ The first was from Dr. Nowak from N. Copernicus Univ. (present address: wieslaw@enstay.ensta.fr). He suggested a paper by Kato and Amamatsu in J.Phys. Chem, 1989 (1990?) (exact reference unavailable) "In this paper a theoretical model of solvation of DMABN is disscused, but real MD simulations are only announced." Also suggested was G.K.Schenter and C.B.Duke, Chem. Phys. Letters 176(1991)563 Theory of photoinduced twisting dynamics in polar solvents: application to DMABN in propanol at low temperatures. ++++++++++++++ A second reply came from Brian Williams at Bucknell Univ. (williams@coral.bucknell.edu) who wrote that he is working on some phenoxazone compounds whose structures are somewhat similar to those showing TICT behavior, and demonstrating solvatochromism in their absorption and fluorescence spectra. In subsequent discussion he suggested a couple of references on TICT and PRODAN including Parassasi et. al. Biophys. J. 57 1179 (1990) and Zurawsky and Scarlata, J. Phys Chem. 96 6012 (1992), as well as work by Rettig, and by Gratton and coworkers. ++++++++++++++ I received a third reply from Don Williams University of Louisville, Kentucky on a program, PDM93 that could be used to perform such calculations. From: WILLIAMS%XRAY2@ULKYVX.LOUISVILLE.EDU PDM93 is a state-of-the-art tool which may be used to model the molecular electrostatic potential for molecular docking calculations. This program may be used to investigate the validity of atomic charge or site multipole models. I append a brief description of this program. -Don Williams Program PDM93, Potential Derived Multipoles The following is a brief description of this program. Molecules interact with each other via their electric potential. PDM93 finds optimized net atomic charges and other site multipole representations of the molecular electric potential based on a variety of models. The program is easy to use, flexible and powerful. Results are obtained in a single iteration and a complete error treatment is made which includes estimated standard deviation and correlation of variables. The program is written in Fortran 77 and runs on Unix, Vax, and other computers with F77 capability. (stuff deleted) The rest of the note went on to describe some features and ordering info. ----------------------- Finally, I decided to compile a list of additional references on the subject which I include here. This is a partial list at best, and I encourage readers to send any additions to this list to me. I will of course compile these and send them out if there is sufficient interest. Thanks to all who replied, or may do so in the future. References on computational chemical approaches to compounds with TICT involving rotation about aryl-NR2 bonds (so far): Rettig,W., Bonacic-Koutecky,V. (1979) Chem Phys Lett 62, 115 Lipinski,j., et al. (1980) Chem Phys Lett 70, 449-453 Bonacic-Koutecky,V Michl, J., (1985) J Am Chem Soc 107, 1765-1766 Nowak, W, et al. (1986) J. Molec. Struct 139, 13-23 Rulliere, C., et al. (1988) Chem Phys Lett 143, 565-570 Illich,P., Pendergast,F.G., (1988) J Phys Chem 93, 4441-4447 Reviews on TICT: Grabowski, Z.R., et al. (1979) Nouv. J Chim. 3, 443-454 Rettig,W., (1986) Angew Chem Int Ed Eng 25, 971-988 ------------------------------------------------------------------------- From: U6124673@ucsvc.ucs.unimelb.edu.au Dear Jon, You may find the following references useful (& refs therein): S. Larsson, E-T in Proteins, J.Chem.Soc., Faraday Trans., 79, 1375 (1983); S. Larsson & A. Volosov, Dist Dependence in photo-induced intramol E-T, J.Chem.Phys., 85, 2548 (1986); Ibid, 87, 6623, (1987); S. Larsson & B. Kallbring, Charge sepn in 9-9'-dianthryl & a special pair treated by a semi-empirical reaction field method, Int.J.Quant.Chem.:Quant.Biol.Symp., 17, 189 (1990); A.H.A. Clayton, et.al., A semi-empirical investigation of C-T interactions in rigid dimethoxynaphthalene-bridge-pyridinium systems, J.Phys.Chem. (1993) - a recent edition, I'm not sure which one. As far as I understand it these refs use CNDO/S/CI method. The C-T state is "created" by varing an electric field to simulate solvent reorganization. Hope you find something useful, regards ------------------------------------------------------------------------- From: wieslaw@enstay.ensta.fr (chercheur-AP) Dear Dr Jacobson, I have noticed that in your posting on MD calculations for TICT related molecules the journal with the paper by Kato and Amamatsu is erronously indicated. The full reference is: S.Kato and Y.Amamatu, J.Chem.Phys.,92,7241 (1990). (I am not sure spelling of the last author! I have no paper nor journal with me!). You might be interested also in the other paper by S.Kato: "Dielectric relaxation dynamics os water and methanol solutions associated with the ionization of N,N-dimethylaniline: Theoretical analyses." K.Ando, S.Kato, J.Chem.Phys. 95,5966 (1991). Numerous references to theoretical works (quantum - chemical) on TICT exhibitig molecules may be also find in our paper "Theoretical study of singlet excited states of donor-acceptor naphthalene derivatives" W.Nowak and W.Rettig , will apear in J.Mol.Struct. (Theochem), probably this year (galey proofs were send back in May). You may be also interested in experimental works on excited state potentials, for example: "Conformational changes upon S1<--S0 excitation in 4-DMABN and some of its chemical analogs" , by V.H.Grassian,J.A. Warren,E.R.Bernstein and H.V.Secor. J.Chem.Phys. 90,3994 (1989). ------------------------------------------------------------------------- From: elewars@alchemy.chem.utoronto.ca (E. Lewars) There is a review on MINDO/3 in Chem Rev, ca. 1984, which likely has refs to charge-transf complexes among its numerous refs. Michael Dewar, the main developer of semiempirical methods, discusses charge-transfer complexes briefly on ps 121-122 of his autobiography, A Semiempirical Life (available from the American Chem Soc); the refs he gives are JACS 1964 1592 Tetrahedron Suppl 1966 7 97 Tetrahedron Letters 1967 5043 JACS 1968 90 1075. ------------------------------------------------------------------------- From: PA13808%UTKVM1.BITNET@uga.cc.uga.edu It is difficult to answer your question without knowing more about the type of "CHARGE TRANSFER" complex your are interested in.If you are concerned with geometry it is probably better to forget the term charge transfer and to think of them as weak complexes held together by dispersion forces and dipole-dipole forces(MULTIPOLE-MULTIPOLE and induced multipole multipole forces really). Then you have a whole r of models available.Very few "CT complexes"involve much charge transfe r in the ground state, The CT excitation which characterises them is to an excited state in which an electron is transferred from a donor to an acceptor. For example many PI/PI* type complexes. The only ones which involve CT in ground state are those in which a lone pair interacts with a polarisable acceptor eg an amine complexing with the iodine molecule.(Characterised by a considerable dipole moment change in the ground state. General refs Theory of Molecular Interactions by I.G.KAPLAN(1986) Theoretical Models of chemical Bonding ed Z.B.Maksic PART 4 Theoretical Treatment of Large Molecules and Their Interactions (Springer Verlag 1991)see also part3 Molecular Spectroscopy, Electronic Structure and Intramolecular Interactions. (Springer Verlag1991 John E. Bloor (PA13808 at UTKVM1.UTK.EDU) ------------------------------------------------------------------------- -- ################################################################ # Jon Erickson e-mail: jaeric@mtcamm.monsanto.com # # Monsanto Company, U3E phone: (314) 694-1511 # # 800 N. Lindbergh Blvd. FAX: (314) 694-1080 # # St. Louis MO, 63167 # # # # What is a magician but a practising theorist? # # -- Obi-Wan Kenobi # ################################################################ From marty@ionchannel.med.harvard.edu Fri Oct 1 10:26:48 1993 Date: Fri, 1 Oct 93 17:26:48 -0700 From: marty@ionchannel.med.harvard.edu (Marty Gallagher) Message-Id: <9310020026.AA05238@ionchannel.med.harvard.edu> To: CHEMISTRY@ccl.net Subject: torsion of conjugated systems -- summary Hello, I am posting a compilation of all the responses I got concerning the question I had posed: |> Does anyone know a forcefield which includes a term for |> torsional energy in a conjugated system? Specifically, I am |> interested in rotating the carbonyl in acetophenone out of plane |> with the phenyl ring. MOPAC geometry optimization places the |> carbonyl in the plane of the ring. Without running a dozen |> MOPAC runs, can I get an estimate of what the energy would be |> out of the plane? To give some background on the problem: I had originally used one molecular mechanics program (which I'll leave nameless because its failure could easily have been due to my own naivete rather than lack of parameterization) which said the lowest confomer was 90 degrees! Since last week when I've tried it again using a different MM package, (biosym's discover), the lowest confomer is at 0 degrees which is 10 kcal lower than the confomer at 90 degrees. The phi potential accounted for almost 100% of the energy difference Hopefully, this will result will be similar to the MOPAC run which I am performing now. The reason I hope that I can perform these calculations via MM rather than QM is because I am really not interested in acetophenone, but in compounds that that have that conjugated system plus 5-8 other interesting dihedral angles. I realized that searching enough of conformational space with these many torsional angles is prohibitive enough without throwing in any QM calculations. Thank you, everyone who wrote! I learned quite a bit. =================================================================== | | | Martin J. Gallagher | | Dept of Neurobiology | | Harvard Medical School | | 220 Longwood Ave | | Boston, MA 02115 | | (617) 432-1729 | | | | marty@ionchannel.med.harvard.edu | | | =================================================================== responses: ================================================================================ Unfortunately, I can not really help you here. An arguement in favor of doing the dozens of MOPAC runs: there will be significant electronic state shifts as you distort the conjugated system likely leading to significant changes in charge distribution. No empirical force field is going to get these right. Electrostatics and hydrogen bonding are critical deteriminants of specificity in many active sites, and Ray Salemme argues convincingly that most really tight ligand/protein interactions involve some sort of "special" chemistry such as cooperative hydrogen bonding. With some semiempirical data, you at least have a shot at getting on handle on these. David States, Director, M.D., Ph.D. '83 (Harvard Med/Biophysics) Institute for Biomedical Computing Washington University in St. Louis Pieter Stouten writes, Conjugated pi-systems constitute one of the larger issues in the development of force fields that describe macromolecule-drug interactions adequately. Anyone interested in this may like to contact M.J.Bearpark, M Olivucci or M.A.Robb who gave a lecture on "Large active space valence bond methods: application to conjugated polycyclic hydrocarbons" This theory takes an MM sigma framework then uses ab-initio parametering on top to modify the force constants, I could not follow most of the technical side of how this was done but the results looking promising. Email address: max@gandalf.ciam.unibo.it (Olivucci) udca700@oak.cc.kcl.ac.uk (M.A. Robb) no email for Bearpark. Andy. Martin, The problem you pose, as I understand it is - How do you determine the energy of a conjugated acetophenone with the acetyl group tilted out-of-plane to various degrees? With that in mind, what you need to use it the "dihedral angle driver" that is present in either MOPAC (MNDO or AM1), or in a variety of molecular mechanics programs such as MM2, etc. The principle involved is simply rotating the selected dihedral angle through the desired arc, using the desired increment, and doing without minimization, or geometry optimization) an energy calculation at each point. I have in the past used SYBYL SEARCH, and MacroModel to do this. I believe MNDO also has a similar routine, but I have not used it. Additionally, SYBYL or AMBER will do the rotation, hold the rotatable angle fixed, and minimize the rest of the molecule, which is also a very good idea. Molecular mechanics should be as good, or better than, a semi-empirical method for the shape of a molecule for which it is well parameterized. It should also give good relative energies between different conformations of the same molecule, and between similar molecules. Molecular mechanics also has the virtue of being faster. If I can be of any further help let me know. I love your library on Longwood. I was a chemistry graduate student at MIT (about 20 years ago) and had the opportunity to use the library there, since my thesis involved mimicry of some features of biological catalysis,and some of the literature in which I was interested was not available at the Institute. Good luck, and best wishes, Dick ************************************ * Richard W. Harper, Ph.D. * * Eli Lilly & Company * * Lilly Research Laboratories * * Lilly Corporate Center * * Indianapolis, IN 46285-0444 * ************************************ * (317) 276-5990 Voice * * (317) 276-5187 Fax * * EMail HARPER_RICHARD_W@lilly.com * * or RWHARPER@acm.org * * Compuserve 73277,1777 * ************************************ From: HARPER RICHARD W (MCVAX0::WALTZ) To: FOREIGN TRANSPORT ADDRESSEE (MCDEV1::IN%"marty@ionchannel.med.harvard.edu") cc: HARPER RICHARD W (MCVAX0::WALTZ) You could do a literature search (CAS on-line) to find out if it's been done. But if it hasn't, what's wrong with a dozen MOPAC runs? You could do it as a reaction coordinate with 1 input file. If you left it on a VAX Friday night, it would probably be done Monday AM... Irene Newhouse Marty, My advice to you is to perform your calculations using the HF/6-31G* theoretical model within a program like Gaussian9x. The energies that you are interested in examining are going to be quite sensitive to the methods used in determining them, so it pays to use a certain minimum level of theory which has been proven to work in cases like this. I realize that you may not want to commit the time and resources to an ab initio study, but you will get much better results if you do. Curt Breneman Asst. Professor of Chemistry Rensselaer Polytechnic Institute Troy, NY 12180 breneman@quant.chem.rpi.edu (and others) In some recent postings to the Computational Chemistry List the issue of force fields for conjugated pi-systems was brought up. Marty Gallagher wrote: > Does anyone know a forcefield which includes a term for > torsional energy in a conjugated system? > Conjugated pi-systems constitute one of the larger issues in the development of force fields that describe macromolecule-drug interactions adequately. Efforts generally seem aimed at sort of ad hoc development of parameters for a small number of closely related compounds. Often times these will not be transferable and consistent with parameters that have been developed in the same vein for other compounds containing the same functional groups. We are in desparate need of a concerted effort towards a consistent treatment of conjugated pi-systems. I sincerely hope this will be a (the ?) main focus of the Biosym Potential Energy Forcefield Consortium over the next years. I also hope that more institutions will join the consortium so that the development of the force field can be given some more momentum. Pieter Stouten, Senior Research Scientist || Computer Aided Drug Design Group || The Du Pont Merck Pharmaceutical Company || Adventures get spoiled P.O. Box 80353, Wilmington, DE 19880-0353 || by being reduced to data Phone: +1 (302) 695 3515 || -- Fax: +1 (302) 695 2813 || Poul Anderson ARA/Fax: +1 (302) 695 4324 || E-mail: stoutepf@chemsci1.es.dupont.com || --- 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 CHEMISTRY-SEARCH@ccl.net --- search the archives, read help.search file first --- On Wed, 29 Sep 1993, Marty Gallagher wrote: > Does anyone know a forcefield which includes a term for > torsional energy in a conjugated system? ... Hi! You may find that most molecular mechanics programs include such terms. Programs that you may consider: Peter Kollman's Amber Allinger's MM2 series (look at MM3 I think they have addressed a problem similar to yours) CharmM (Chemistry Dept. at Harvard) Gromos Also, check out papers by Beveridge published possibly in JACS in the '60's - '70's. He did a torsional analysis on acetylcholine and you may get ideas on how to proceed with your work. gus mercier mercie@cumc.cornell.edu ----------------------------- Application message id: STLEDD 930929104214605011 Importance: Normal Grade of Delivery: Normal ----------------------------- Martin; Just set the dihedral angle of interest to whatever value you wish, such as 90 degrees. Also, change the optimization flag from one to zero for the dihedral you just changed. MOPAC will optimize the geometry to at least a stationary point (not necessarily a minimum) without rotating the dihedral angle. If you are using the QCPE Draw program to create the input data set this will be easy. Other packages will also provide ways to "fix" an geometry spec. so that it is not adjusted during optimization. You'll have to check the manual for the program you are using. Alternatively you could go to the operating system, use the editor and make the changes you need (This assumes you can identify the atoms in the data file.) Good Luck, Fred Just for fun I constructed acetophenone and ran a dihedral driver 0 to 90 degrees using the molecular mechanics program MM2(91). I used my own molecular modeling system, MacMimic, that runs on Macintosh II and Quadra computers, and I had the results within three minutes. Here is the output file from the driver. The energies are in kcal/mol. I would expect the MM2 energies should be of very high quality (equal to or better than MOPAC) in this case. acetophenone ANGLE1= 2- 3- 12- 14 0.0 0.3684 5.0 0.3751 10.0 0.4042 15.0 0.4587 20.0 0.5440 25.0 0.6647 30.0 0.8276 35.0 1.0336 40.0 1.2807 45.0 1.5633 50.0 1.8680 55.0 2.1868 60.0 2.5022 65.0 2.7946 70.0 3.0460 75.0 3.2397 80.0 3.3588 85.0 3.3918 90.0 3.3326 TOTAL ELAPSED TIME IS 123.18 SEC. -Anders -- Anders Sundin e-mail: Anders.Sundin@orgk2.lth.se Organic Chemistry 2 ok2aps@selund.bitnet Lund University, P.O. Box 124 voice: +46 46 104130 S-22100 Lund, Sweden fax: +46 46 108209 You could look at the torsional potential in MMP2. Better yet, do the calculations at different torsional & compare the energies. Yvonne Martin, Senior Project Leader @@@@@@@@@@ Computer Assisted Molecular Design Project @ D-47E, AP9A-LL @@@@@@@@ Abbott Laboratories @ @ One Abbott Park Road @ @ Abbott Park, IL 60064 @@@@@@@@@@ Phone: 708 937-5362 FAX: 708 937-2625 Internet: MARTINY@ABBOTT.COM The way e.g. MMP2 handles this is to do 'pi-calculations' prior to the molecular mechanics energy minimization. From the 'pi-calcs' one get the bond order of the pi bonds. The bond orders are used to scale the bond force constants and the torsional barriers for the given pi bonds. Depending on the amount of rotation from 'normal', the bond order and hence the parameters are changed. I guess others have used this or similar approach. You may take a look at Burkert & Allinger: Molecular mechanics, ACS monograph 177 (1982), p. 52 and on... Disclaimer: I haven't used MMP2 for several years ... -oed. --- 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 CHEMISTRY-SEARCH@ccl.net --- search the archives, read help.search file first --- Greetings, I'm assuming you're talking about molecular mechanics, and not ab initio or sem-empirical level calculations. Either of the latter would automatically consider conjugation, and although ab initio calculations are considered "better", they require a lot more computing power and without a supercomputer are unsuitable for even moderate sized molecule. Semi empirical calcs, however, can be carried out on a 486, mini, micro or non-parallel computer for even fairly large molecules. Programs of this type include MOPAC, AMPAC (MNDO, MNDO/3, AM1) and are available both commercially and from QPCE (Quantum Chemistry Program Exchange at Indiana University). As for molecular mechanics, some of the programs have built in conjugation terms. The one organic chemists use most is MMX. Hope this helps Kimberley Cousins Cal State Univ San Bernardino kcousins@wiley.csusb.edu Force fields that include explcit conjugation effects in determining K(stretch), L0, and V2 have been extremely valuable here at Kodak for determining dihedral angles of extensively conjugated systems, namely MMX. This is the forcefield found in in PCMODEL from Serena Software, and available also in MODEL, from Kosta Steliou, now in Boston, I believe. MMX "nails" the dihedrals of biphenyl, o-Me biphenyl, and cisoid o,o'-dime biphenyl to within 1-2 degrees of experiment (42,58, and 72 degrees, respectively.) within 1-2 degrees of experiment (42,58, and 72 degrees, respectively.) It will show some effect on the torsion angle of biphenyl when substituted 4-no2,4'-nh2. It is easy to "tune the forcefiled if desired...it is editable in a flat file. Hope this is useful, inspite of missspelings. John McKelvey --- 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 CHEMISTRY-SEARCH@ccl.net --- search the archives, read help.search file first --- Marty, Though I don't have an in depth familiarity with MOPAC, I am fairly well versed in QM/MM and modelling techniques. So I would assume that the following is applicable to MOPAC. When you set up your z-matrix defining the torsion for your carbonyl, it is possible to scan the potential energy surface for the torsional space that you have defined for the carbonyl. In gaussian92 the keyword is "scan". I would assume MOPAC has something similar. You should be able (with that keyword) to define a variable - in this case your torsion, define the increment value you would like (e.g. say you wanted the dihedral to vary by 30 degrees each step) then add the number of steps you would like (obviously, you can use symmetry to cut down on the number of runs to generate a PE profile). Also as semi-empirical methods are rather quick on most computers, you should get a very quick answer to your problem. Hope this helps! ********************************************** * * * Mark A. Zottola * * markz@chem.duke.edu * * Department of Chemistry * * Duke University * * * * 'The fault, dear Brutus, lies not with * * ourselves, rather within our CPU's...' * * (with due apologies to W.S.) * * * ********************************************** Generate one starting structure with the carbonyl out of the plane of the ring, convert that file to a MOPAC file format, and perform the calculation with 1SCF in the first line. This will not optimize the geometry, but will calculate a single point energy for the input structure. If you don't have a way to generate starting structures of choice, let me know. Abby Parrill abby@mercury.aichem.arizona.edu The University of Arizona-chemistry I believe the MM2(or MM3) forcefield would handle this situation quite well. It includes a dihedral driver routine which is automated. This will help you locate the transition state of the rotation, whose energy will give you the barrier. This may have been done for acetophenone, so I would suggest a quick literature search would save you some time. You probably realize this, but from the wording of the question I thought I'd clarify. MOPAC is a semi-empirical molecular orbital method, that is parameterized, but not in the sense of a forcefield, ie. bond, angle, and torsion terms. If you are using MOPAC for a torsion analysis, you can step through the angles by placing a -1 in the column following the dihedral angle of interest, and then put all the steps on a line at the end of the file. If you really want just a quick estimate, just manually set the dihedral to be out of the ring and do a single point energy calculation. Just be aware that this geometry may not be a minima. -- ################################################################ # Jon Erickson e-mail: jaeric@mtcamm.monsanto.com # # Monsanto Company, U3E phone: (314) 694-1511 # # 800 N. Lindbergh Blvd. FAX: (314) 694-1080 # # St. Louis MO, 63167 # # # # What is a magician but a practising theorist? # # -- Obi-Wan Kenobi # ################################################################ --- 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 CHEMISTRY-SEARCH@ccl.net --- search the archives, read help.search file first --- On the assumption that it is cpu time you wish to save rather than the number of runs, you might try PCMODEL, which has a dynamically determined conjugated force field...built off Allingers MP1 method. We use it extensively for dyes here at KODAK and the MMX forcefield in PCMODEL saves us tons of time, and it runs on pc's, mac,you-namit workstations. Available from Kevin Gilbert 812-855 1302 or Joe Gajewski 885-1192. John McKelvet Dear Mr. Gallagher: Reliable net atomic charges can be found by fitting the molecular electric potential obtained by ab initio quantum mechanical methods. Program PDM93 also allows atomic dipoles/quadrupoles, bond dipoles, and addition of lone pair electron sites if necessary in order to fit the electric potential. A particularly useful feature of the program is the transparent way in which dependency conditions are specified. I append information about this program. -Don Williams Program PDM93, Potential Derived Multipoles The following is a brief description of this program. Molecules interact with each other via their electric potential. PDM93 finds optimized net atomic charges and other site multipole representations of the molecular electric potential based on a variety of models. The program is easy to use, flexible and powerful. Results are obtained in a single iteration and a complete error treatment is made which includes estimated standard deviation and correlation of variables. The program is written in Fortran 77 and runs on Unix, Vax, and other computers with F77 capability. Program PDM93 has a unique combination of features: o general sites, not necessarily at atomic locations o each site may have any combination of monopole, dipole, or quadrupole o bond dipole model is supported o restricted (along the bond direction) bond dipole model is supported o provision for site dipole vectors in sp2 or sp3 directions o selected fixed atomic charges o selected groups of atoms with fixed charge o atomic charge equalities or symmetry relations o rotational invariance of site charges o provision for optional foreshortening of X-H bonds o comparison with Mulliken charges and Mulliken electric potential o direct input from Gaussian-92 o generalized input from other quantum mechanics programs o automatic generation of electric potential grid points o provision for custom generation of grid points o on-line program manual o comprehensive examples are provided A review of potential-derived charges may be found in Reviews of Computational Chemistry, Vol. II, pp. 219-271 (1991). For further information contact Dr. Donald E. Williams, Department of Chemistry, University of Louisville, Louisville, Kentucky 40292, USA. Tel:(502)588-5975 Fax:(502)588-8149 E-mail:dewill01@ulkyvx.louisville.edu ----------------------------------------------------------------------- Ordering information Program package consisting of manuals, Fortran-77 source files, and demonstration example files....................................$2,000 Special discount price is available to academic institutions ........$495 Choose the preferred method of shipment: Via ftp, purchaser furnishes valid ftp address................n/c Via magnetic media, specify type..............................$20 DC6150 tape cartridge, unix tar format or MP120 tape cartridge (8mm), Vax backup format or 9-track 1600 bpi ascii tape or 9-track 6250 bpi ascii tape Make check payable to the University of Louisville. ------------------------------------------------------------------------- Dear Marty, To calculate the torsional energy barrier of a system like acetophenone with semiempirical or ab initio methods is not so difficult since the low energy rotational isomer probably (however see below) has a c-c-C=O torsional angle = 0 deg (where c = aromatic carbon and C = non aromatic carbon atom) and the high energy rotational isomer is almost certainly at 90 degrees (perpendicular rotational isomer). In principle, one only needs to geometry optimize both rotational isomers (holding the c-c-C-C torsional angle fixed of course in the Z-matrix). Then one simply takes the difference in energy of these two conformers as the torsional energy barrier. If the minimum is not at zero degrees and/or the maximum is not at 90 degrees, it gets more complicated. To be on the safe side, it is better to calculate the energies at torsional angles of 0, 15, 30, 60, 75, and 90 degrees and then plot energy versus the torsional angle (by symmetry one also has the -30, -15, 105, 120, 150, 165, and 180 torsional energies and these can be included in the plot). A spline fit of the plot using one of many statistical graphics packages for PC's or Mac will produce a curve. The top of the curve should be at 90 degrees and the bottom at 0 degrees. If either appears displaced from these values, take the apparent minimum or maximum from the plot, set the torsional angle of your starting geometry to that value, and do a complete energy minimization in the case of the torsional minimum or a "transition-state" optimization in case of the torsional maximum. Again, for simple systems like acetophenone, it is usually safe to assume that the maximum is at 90 degrees. The minumum however may be displaced from zero degrees. To check this possibility it is best to displace the structure slightly from zero degrees and minimize. The reason for this displacement and extra mimimization is that many minimizers get "stuck" on planar structures even though the minima may be non-planar. This applies to all techniques, molecular mechanics included. If you would like to use molecular mechanics to examine this system, and one wants to do a careful job, one should use a method like MMP2 which will modify the bond order of a conjugated system as a function of the torsional angle. Finally, most standard molecular mechanics methods like MM2 contained in MacroModel 4.0 (Clark Still, Columbia University) generally don't do too bad a job in estimating these torsional barriers, since MacroModel MM2 at least has been parameterized for these types of systems. MacroModel produces a barrier of 5.81 kcal/mole for acetophenone which seems reasonable. If one wants accurate estimates of these barriers, one should however use high level ab initio calculations (6-31G* at least). Finally, I am somewhat interested in your modeling of the nictotinic acetyl choline receptor since we have previously developed a pharmacophore model of the muscrinic M2 receptor and I have a long standing interest in ion channels. In addition, Lynn Jelinski of Bell Labs studied the binding of acetyl choline to the nictotinic receptor using transfer NOE techniques where the free acetyl choline retains a memory of its bound conformation. However when we analyzed the conformation they claimed as being the active one, it was so high in energy, it was implausible. Someone should really go back and re-exaine that data with a careful modeling study. One additional point: We have developed some software called APOLLO (1-4) (Automated PharmacOphore Location through Ligand Overlap) which might be useful for your modeling study. Basically the program searches for low energy conformers of ligands which allow good overlap of pharmacophore point and therefore would "explain" the binding of a set of ligands to a common receptor. We have already used this software to develope pharmacophore models/active conformation hypotheses for the muscarinic receptor (1), NMDA agonists and antagonists (2-4), as well as the benzodiazepine receptor (5). The software works best in conjunction with MacroModel. If you have MacroModel you might be interested in using our program. If so, let me know and I will send you a copy. It currently runs on a VAX, but I am in the midst of porting it to Unix. The port should be completed in a couple of weeks. Sincerely, ------------------------------------------------------------------ | Konrad Koehler | Computational Chemistry Group | | internet: koehler@irbm.it | Department of Medicinal Chemistry | | | IRBM | | telephone: +39-6-910-93606 | Via Pontina Km 30.600 | | fax: +39-6-910-93225 | 00040 Pomezia (Roma) | | | Italy | ------------------------------------------------------------------ (1) Koehler, K. F.; Spangler, D. P.; Snyder, J. P. Pharmacophore Identification through Molecular Similarity. Division of Computers in Chemistry Computational Graph Theory and Combinatorics; Molecular Similarity Poster Session Tuesday, September 27 Poster #35. 196th American Chemical Society National Meeting, September 27; American Chemical Society: Los Angeles, CA, 1988. (2) Snyder, J. P.; Koehler, K. F.; Spangler, D. P. Searle R & D Group Employs Receptor Mapping as a Prelude to Drug Design. Chem. Design Auto. News 1989, 4, 1. (3) Snyder, J. P.; Rao, S. N.; Koehler, K. F.; Pellicciari, R. Drug Modeling at Cell Membrane Receptors:the Concept of Pseudoreceptors In Trends in Receptor Research; Angel, P.; Gulini, U. and Quagli, W., Eds.; Elsevier: Amerdam, 1992; pp 367-403. (4) Snyder, J. P.; Rao, S. N.; Koehler, K. F.; Vedani, A.; Pellicciari, R. APOLLO Pharmacophores and the Pseudoreceptor Concept In Trends in QSAR and Molecular Modeling; Wermuth, C. G. and Rival, Y., Eds.; Elsevier: Amsterdam, 1993; pp in press. (5) Diaz-Arauzo, H.; Koehler, K. F.; Hagen, T. J.; Cook, J. M. Synthetic and Computer Assisted Analysis of the Pharmacophore for Agonists at Benzodiazepine Receptors. Life Sciences 1991, 49, 207-216. To answer the request for dihedral angle energy barriers: The following reference is a good place to start. It will lead you many other references with respect to dihedral angle energy barriers of MM2 as compaired to variable temp. NMR studies for small amines. Brown, J. H.; Bushweller, C. H.; J. Am. Chem. Soc. 1992, 114, 8153 --- 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 CHEMISTRY-SEARCH@ccl.net --- search the archives, read help.search file first From fredvc@esvax.dnet.dupont.com Fri Oct 1 12:29:56 1993 Date: Fri, 1 Oct 93 16:29:56 -0400 Message-Id: <9310012029.AA07901@esds01.es.dupont.com> From: fredvc@esvax.dnet.dupont.com To: chem@ccl.net Subject: THE PARAMETRIC NATURE OF VDW RADII Chris Cramer writes: >> van der Waals radii are not physical constants, they are parameters. >> >> I make this point because I see a lot of modeling done which involves >>taking the "van der Waals surface" and evaluating some property there. While >>this particular property may have nothing to do with the physical >>measurements which were used to develop the vdW radii from some OTHER >>physical model, nevertheless they are justified as somehow being "better" >>than parameters which are in fact optimized for EXACTLY the property being >>studied. If anyone believes that VDW radii are "real", I urge them to take a look at the original MMI paper [JACS, v89, p4345 (1967)] Reasonalbe results were ob- tained with a carbon vdw radius of ~1.7 Ang,....... and ~1.1 Ang!! ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ FREDERIC A. VAN-CATLEDGE Scientific Computing Division || Office: (302) 695-1187 or 529-2076 Central Research & Development Dept. || The DuPont Company || FAX: (302) 695-9658 P. O. Box 80320 || Wilmington DE 19880-0320 || Internet: fredvc@esvax.dnet.dupont.com -------------------------------------------------------------------------------- Opinions expressed in this electronic message should ***> NOT <*** be taken to represent the official position(s) of the DuPont Company. *****> ANY OPINIONS EXPRESSED ARE THE PERSONAL VIEWS OF THE AUTHOR ONLY. <***** ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From jomal@chopin.dfq.ufrj.br Fri Oct 1 18:07:21 1993 Date: Fri, 1 Oct 1993 18:07:21 GMT From: jomal@dfq.ufrj.br (Joao Otavio M.A. Lins ) Message-Id: <9310011807.AA16706@dfq.ufrj.br> To: chemistry@ccl.net Subject: Iron complexes w/ MM Hi Netters! A friend of mine has asked me about the possibility to use molecular mechanics to perform calculations on iron complexes. As a second opinion, I'd like to hear your point of view as well as where he'll find parameters/ force fields available to perform such calculations. Are they reliable? Please, send replies direct to me and I'll summarize to the net. Thanks in advance, Joao O.M.A. Lins Theoretical Chemistry Group ______________________________________________________________________ Instituto de Quimica da U.F.R.J. email: jomal@dfq.ufrj.br Centro de Tecnologia, bloco A, sala 412 voice: (55)(21) 590-9890 Cidade Universitaria fax : (55)(21) 290-4746 21949-900 - Rio de Janeiro - RJ BRAZIL ______________________________________________________________________ From hogue@calumet.den.mmc.com Fri Oct 1 12:07:30 1993 Date: Fri, 1 Oct 93 18:07:30 MDT From: hogue@calumet.den.mmc.com (Pat Hogue 1-2183) Message-Id: <9310020007.AA04294@calumet.den.mmc.com> To: chemistry@ccl.net Subject: van der Waals radii Dear Netters: Is is logical to think of the van der Waals surface of organic molecules as the analog of the solid state Fermi surface? Pat Hogue