From rabe@fu-berlin.de Wed Nov 19 04:34:04 1997 Received: from mail.chemie.fu-berlin.de for rabe@fu-berlin.de by www.ccl.net (8.8.3/950822.1) id DAA04133; Wed, 19 Nov 1997 03:58:05 -0500 (EST) Received: from Fermi.Chemie.FU-Berlin.DE(fermi.chemie.fu-berlin.de[160.45.26.7]) (2514 bytes) by mail.chemie.fu-berlin.de via smail with P:smtp/R:bind_hosts/T:inet_zone_bind_smtp (sender: ) id for ; Wed, 19 Nov 1997 09:58:05 +0100 (MET) (Smail-3.2.0.98 1997-Oct-16 #25 built 1997-Nov-07) Received: by Fermi.Chemie.FU-Berlin.DE (Smail3.2.0.98) from Fermi.Chemie.FU-Berlin.DE (127.0.0.1) with smtp id ; Wed, 19 Nov 1997 09:57:54 +0100 (MET) Sender: rabe@www.ccl.net Message-ID: <3472AA11.2847@fu-berlin.de> Date: Wed, 19 Nov 1997 09:57:53 +0100 From: Bjoern Rabenstein Organization: Freie Universitaet Berlin, Germany X-Mailer: Mozilla 3.0 (X11; I; IRIX 5.3 IP22) MIME-Version: 1.0 To: Computational Chemistry Mailing List Subject: G94 - CHELPG-Parameter: summary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Thanks for the useful replies I got. Here the results: (i) Setting VDW-radii: The option you need is POP=(ChELPG,ReadRadii) and then after the structure input you place a block of lines atom-type radius for each element you care to define and end it with a blank line. Example: #P RHF/GEN Pop=(CHELPG,ReadRadii) MaxDisk=2000MB ZnH2O for charges 2 1 O 9.957 29.927 22.183 H 10.538 29.398 22.729 H 10.392 30.778 22.124 ZN 7.852 29.152 22.941 Basis set specifications (blank line) Zn 2.0 ! Zn vdw radius read in here (Thanks to Lars Hemmingsen (nice User-ID, it's the name of my brother... :-) and Doug Fox .) (ii) Modifying the grid: It seems that such a modification needs recompilation of the code (routine BGrid in g94/l602.F according to Doug Fox). The following came from Ulf Ryde (marvellous, Ulf is the name of my father... :) : > You can affect the ESP points by > IOp(6/41=number_of_layers) > IOp(6/42=density_of_points) > but I do not know exactly how they affect the CHelpG points. > The options are not very well documented, so you probably have > to check the code directly. > IOp(6/33=2) is useful to see exactly where the ESP is calculated. I think, "number_of_layers" sounds much like CHELP and not CHELPG. Bye! -- Bjoern Rabenstein * PhD student * Freie Universitaet Berlin Inst. f. Kristallographie * AG Knapp * Takustrasse 6 * D-14195 Berlin [phone] +49-30-838-3484 [fax] +49-30-838-3464 [email] rabe@fu-berlin.de [WWW] http://www.chemie.fu-berlin.de/~rabe/ From destack@unomaha.edu Wed Nov 19 10:34:32 1997 Received: from s-cwis.unomaha.edu for destack@unomaha.edu by www.ccl.net (8.8.3/950822.1) id JAA05507; Wed, 19 Nov 1997 09:52:01 -0500 (EST) Received: from cas.unomaha.edu (cas.unomaha.edu [137.48.34.40]) by s-cwis.unomaha.edu (8.8.5/8.8.5) with ESMTP id IAA05215 for ; Wed, 19 Nov 1997 08:48:03 -0600 (CST) Received: from CAS/MAILQUEUE by cas.unomaha.edu (Mercury 1.21); 19 Nov 97 08:52:04 -600 Received: from MAILQUEUE by CAS (Mercury 1.21); 19 Nov 97 08:51:42 -600 Received: from zinc.unomaha.edu by cas.unomaha.edu (Mercury 1.21); 19 Nov 97 08:51:37 -600 Received: by zinc.unomaha.edu with Microsoft Mail id <01BCF4C8.820C9330@zinc.unomaha.edu>; Wed, 19 Nov 1997 08:52:47 -0600 Message-ID: <01BCF4C8.820C9330@zinc.unomaha.edu> From: "Douglas E. Stack" To: "CHEMISTRY@www.ccl.net" Subject: Imaginary Freq. Date: Wed, 19 Nov 1997 08:52:46 -0600 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable I've had a calculation converge to a transition state structure, but the = value of my one imaginary frequency seems to low ca. 7 cm-1. Animation of the mode seems to connect the nucleophile to the Micheal = acceptor, but I was under the impression that the absolute value of the = imaginary freq. should b between 600 to 200 cm-1. I have a ring system = composed of two fused six membered rings (the Micheal acceptor) would = this make for a low value? Douglas E. Stack=09 Assistant Professor Department of Chemistry University of Nebraska at Omaha Omaha, NE 68182-0109 (402) 554-3647 (402) 544-3888 (fax) destack@unomaha.edu From s.hogg@ic.ac.uk Wed Nov 19 11:34:05 1997 Received: from romeo.ic.ac.uk for s.hogg@ic.ac.uk by www.ccl.net (8.8.3/950822.1) id LAA05989; Wed, 19 Nov 1997 11:06:44 -0500 (EST) From: Received: from judy.ic.ac.uk [155.198.5.5] by romeo.ic.ac.uk with esmtp (Exim 1.62 #1) id 0xYCeA-00063j-00; Wed, 19 Nov 1997 16:06:34 +0000 Received: from mtcmsa.mt.ic.ac.uk (root@mtcmsag1.mt.ic.ac.uk [155.198.96.22]) by judy.ic.ac.uk (8.7.5/8.7.5) with SMTP id QAA16192 for ; Wed, 19 Nov 1997 16:06:12 GMT Received: from mtcsi.mt.ic.ac.uk by mtcmsa.mt.ic.ac.uk (5.x/4.1) id AA08791; Wed, 19 Nov 1997 16:06:09 GMT Message-Id: <2977.9711191606@mtcsi.mt.ic.ac.uk> Subject: SOFIE - Any ideas?? To: chemistry@www.ccl.net Date: Wed, 19 Nov 1997 16:06:08 +0000 (GMT) X-Mailer: ELM [version 2.4 PL24] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit This is an email I recevied from a friend (Rob Miller, rmiller@sanbi.ac.za) that may be of interest. If anyone is interested in the project, email me, the list, or Rob directly, and it may be useful. -- Simon Hogg Imperial College, London, UK > Did you see a recent announce on various linux ngs about a > 3d game engine called sofie ? > it's set up so you can define new enemy > spaceships that fly about and respond to their environment... > so I was tinking... any interest in trying to help build a > `chemistry' world for it ? Not QM level simulation of > course, but perhaps enough to get bond angles and structures > basically right -- then you could maybe go and `push' on > an atom and watch the structure deform/respond, or blast > a C atom out of existence and see how the molecule recovers ! > ----- Appended Message ----- > sofie is a real time 3d graphics engine with collision detection system. > A little space fighting game (1 level) as demo is included. > > I already uploaded sofie to sunsite but probably it is > not now in the destination directory. You can also download > sofie from > > http://www.physik.uni-regensburg.de/~scs22156/sofie-0.2/sofie.html > > WANTED: contributors for > > - - creating new space ships with/without textures, or > improving old ones. (no programming skills necessary) > > - - writing AI for other computer enemies (the current are > to dumb). This is an isolated problem, no internals of > sofie are necessary to know. > > - - what else? May be you know ... > From schiffer@h1tw0036.hoechst.com Tue Nov 18 08:33:57 1997 Received: from www.hoechst.com for schiffer@h1tw0036.hoechst.com by www.ccl.net (8.8.3/950822.1) id IAA27555; Tue, 18 Nov 1997 08:31:07 -0500 (EST) Received: by www.hoechst.com id OAA29363; Tue, 18 Nov 1997 14:31:13 +0100 Received: from unknown(134.81.76.36) by www id sma029276; Tue Nov 18 14:30:16 1997 Received: from h1tw0024 by h1tw0036 via SMTP (940816.SGI.8.6.9/940406.SGI.AUTO) id OAA19136; Tue, 18 Nov 1997 14:26:11 +0100 Message-ID: <34719773.167E@h1tw0036.hoechst.com> Date: Tue, 18 Nov 1997 14:26:11 +0100 From: "Dr. Heinz Schiffer" Organization: Hoechst Research & Technology X-Mailer: Mozilla 2.0 (X11; I; IRIX 5.3 IP22) MIME-Version: 1.0 To: Frank Jensen CC: Computational Chemistry List Subject: Re: CCL:HF and EC bond lengths References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Frank Jensen wrote: > > Dear All, > > For 'normal' organic molecules the general trend > is that Hartree-Fock geometries have bond lengths which are > too short (at least when sufficiently large basis sets are used). > Addition of electron correlation increases the bond lengths, > and MP2 with a large basis set is often surprisingly close > to the experimental value. This behavior is often rationalized > by arguing that the Hartree-Fock wave function has the > wrong dissociation limit, and by the fact that electron > correlation in general increases as a function of bond length. > > For metal coordinated species (e.g. ferocene), however, > the observed trend is normally the opposite, i.e. Hartree-Fock > geometries have too long bond lengths and electron correlation > makes them shorter. MP2 normally is quite poor in these cases, > and overshoots the correlation effect significantly. Coupled > cluster method (e.g. CCSD) give quite good geometries, as does > DFT. > > Does anyone have a good rational for this change > in behavior? The arguments for the 'normal' molecules should > be equally valid for the metal coordinates species, but the > observed trend is the opposite. > > Frank > Hi Frank, in my view it's especially the intramolecular dispersion energy (which is always attractive) between the d-electrons of the iron and the pi electrons of the cyclopentadienyl rings and of course between the two rings which is missing in HF calculations of ferrocene and which leads to too long distances. MP2 may be bad, because of the too many electrons involved and because the pi electrons are delocalized. Good DFT geometries surprises me a little bit, because it was shown by Pulay, Becke, and Sprik that also DFT does not contain any dispersion contributions at all. May be, these are LDA or LSDA results, which always leads to overbinding and therefore to too short distances. But only in the limit of a complete basis set. So, the good performance of DFT may be the result of a fortuitious error compensation. If gradient corrected DFT-methods are used, the distances should be too long, as with HF. But with small basis sets...see above. By the way, there is an exception to the 'normal' rule, that HF distances in the basis set limit are too small : Si2H6 and Cl2. For both molecules you get larger bond distances on the HF level of theory if you add more polarization functions (2d, 2d1f, etc.). In Cl2 you have again the dispersion interactions of the lone pairs. For Si2H6 I do not have any convincing explanation. I hope I could help a little bit, Ciao Heinz -- Dr. Heinz Schiffer Phone ++49-69-305-2330 Hoechst Research & Technology Fax ++49-69-305-81162 Scientific Computing, G864 Email schiffer@h1tw0036.hoechst.com 65926 Frankfurt am Main Schiffer@CRT.hoechst.com From strahs@winestock.biomath.nyu.edu Tue Nov 18 13:44:40 1997 Received: from choppin.biomath.nyu.edu for strahs@winestock.biomath.nyu.edu by www.ccl.net (8.8.3/950822.1) id NAA28758; Tue, 18 Nov 1997 13:24:03 -0500 (EST) Received: from localhost by choppin.biomath.nyu.edu via SMTP (950413.SGI.8.6.12/930416.SGI) for id NAA03201; Tue, 18 Nov 1997 13:22:58 -0500 Date: Tue, 18 Nov 1997 13:22:58 -0500 (EST) From: Dan Strahs X-Sender: strahs@choppin.biomath.nyu.edu To: CHEMISTRY@www.ccl.net Subject: visualizing normal modes? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I'm trying to visualize some the normal mode eigenvectors for systems I've been working with. I have a number of options, but none really satisfy me. a) I can use Insight/Biosym/MSI to draw arrows attached to the atom positions; but this is an extraordinarily crude approach. It also requires me to iteratively read in and delete the objects associated with the particular mode I want to consider. b) I have code which I can adopt to generate Postscript stereo views of the atoms with the eigenvectors. But these can't be manipulated in 3-D and it would, in the end, use ALOT of paper. Of course, I could save on the paper by using ghostview to display, but that still leaves me with a fixed visualization. c) I can use Geomview to import the atoms as small spheres and draw cones and cylinders for the eigenvectors. This suffers from the same process of iterative reading as the Biosym/Insight solution. Is there some package (preferably free) which will: 1) Display atoms and eigenvectors in a 3-D, rotatable reference frame; 2) Allow me to import the entire calculated set of normal modes and visualize them individually without progressive reloads; 3) Generate some form of output (i.e. Postscript) Thanks for any help that can be offered in this regard. Dan Strahs, Chemistry Dept., NYU From t.niwa@nippon-shinyaku.co.jp Wed Nov 19 21:34:12 1997 Received: from tkusparc.nippon-shinyaku.co.jp for t.niwa@nippon-shinyaku.co.jp by www.ccl.net (8.8.3/950822.1) id VAA13334; Wed, 19 Nov 1997 21:08:56 -0500 (EST) Received: from c0238.nippon-shinyaku.co.jp ([128.1.9.44]) by tkusparc.nippon-shinyaku.co.jp (SMI-8.6/SHINYAKU-1.005/16/97) with SMTP id LAA05692 for ; Thu, 20 Nov 1997 11:05:57 +0900 Message-Id: <199711200205.LAA05692@tkusparc.nippon-shinyaku.co.jp> X-Sender: z3391@172.16.1.251 X-Mailer: Windows Eudora Pro Version 3.0-J (32) Date: Thu, 20 Nov 1997 11:05:46 +0900 To: chemistry@www.ccl.net From: Tomoko Niwa Subject: logP and solubility for Amino Acids and peptides Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-2022-JP" Dear CCL: Wang Arthur wrote: > We are currently on a project of de novo protein design. A set of scales > which describe the hydrophobicities of the natural amino acids is what we > badly need. > (2) LogP values for the 20 natural amino acids and oligopeptides (up to > buta- or pentapeptides). Please check the following papers: Tomoko Sotomatsu-Niwa, Akio Ogino, Evaluation of the hydrophobic parameters of the amino acid side chains of peptides and their application in QSAR and conformational studies, THEOCHEM (1997), 392, 43-54. We statistically analyzed the experimentally determined 1-octanol/water partition coefficient, log P, of a wide variety of N-acetyl-di- and tripeptide amides and unblocked di- and tripeptides to evaluated the hydrophobic parameters of the amino acid side chains of peptides. For blocked peptides, hydrophobic parameters were defined for amino acids having un-ionizable and ionizable side chains. For unblocked peptides, hydrophobic parameters were evaluated for 12 amino acids with un-ionizable side chains. In order to estimate the hydrophobic parameters for unnatural and ionizable amino acids, a procedure was developed based on the relationship between our hydrophobic parameters and the log P values of model amino acids calculated by the fragment constant method. Use of the hydrophobic parameters in quantitative structure-activity relationship (QSAR) studies of bioactive peptides and conformational studies of proteins have also been presented. (Our method is valid for the log P values for buta- or pentapeptides. ) Abraham, Donald J.; Leo, Albert J., Extension of the fragment method to calculate amino acid zwitterion and side chain partition coefficients, Proteins: Struct., Funct., Genet. (1987), 2(2), 130-52 Leo, A. J., Computer calculation of peptide hydrophobicity, Pharmacochem. Libr. (1991), 16(QSAR: Ration. Approaches Des. Bioact. Compd.), 349-52 Fauchere, Jean Luc; Charton, Marvin; Kier, Lemont B.; Verloop, Arie; Pliska, Vladimir, Amino acid side chain parameters for correlation studies in biology and pharmacology, Int. J. Pept. Protein Res. DATE: 1988 VOLUME: 32 NUMBER: 4 PAGES: 269-78 Akamatsu, Miki; Fujita, Toshio, Quantitative analyses of hydrophobicity of di- to pentapeptides having unionizable side chains with substituent and structural parameters, J. Pharm. Sci. (1992), 81(2), 164-74 Akamatsu, Miki; Okutani, Shinichi; Nakao, Kazuya; Hong, Nam Joo; Fujita, Toshio, Hydrophobicity of N-acetyl-di- and tripeptide amides having unionizable side chains and correlation with substituent and structural parameters, Quant. Struct.-Act. Relat. DATE: 1990 VOLUME: 9 NUMBER: Akamatsu, Miki; Yoshida, Yohji; Nakamura, Hideaki; Asao, Masaaki; Iwamura, Hajime; Fujita, Toshio, Hydrophobicity of di- and tripeptides having unionizable side chains and correlation with substituent and structural parameters, Quant. Struct.-Act. Relat. DATE: 1989 VOLUME: 8 NUMBER: 3 [8] R.-S. Tsai, B. Testa, N. El Tayar, P.-A. Carrupt, J. Chem. Soc. Parkin Trans. 2, (1991), 1797. [9] P. Vallat, P. Gaillard, P.-A. Carrupt, R.-S. Tsai, B. Testa, Helv. Chim. Acta, 78 (1995), 471. [11] W. E. Steinmetz, Quant. Struct.-Act. Relat., 14 (1995), 19. [12] N. El Tayar, H. Karajiannis, H. van de Waterbeemd, Amino Acids, 8 (1995), 125. Best Regards =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Tomoko Sotomatsu Niwa, Ph.D. Research Lab. Nippon Shinyaku Co., Ltd. Nishiohji Hachijo Minami-ku Kyoto, 601 Japan tel 075-321-9047, fax 075-321-9038 t.niwa@nippon-shinyaku.co.jp =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= From jkp@wag.caltech.edu Wed Nov 19 22:34:12 1997 Received: from sgi1 for jkp@wag.caltech.edu by www.ccl.net (8.8.3/950822.1) id VAA13377; Wed, 19 Nov 1997 21:35:18 -0500 (EST) Received: by sgi1 id AA10644 (931110.SGI/IDA-1.4.4); Wed, 19 Nov 97 18:35:14 -0800 Date: Wed, 19 Nov 97 18:35:14 -0800 From: "Jason K. Perry" Message-Id: <9711200235.AA10644@sgi1> To: frj@dou.dk Subject: Re: CCL:HF and EC bond lengths Cc: chemistry@www.ccl.net Dear Frank, In regard to your question about correlation shortening bonds in organometallics in contrast to the situation for organic molecules, I see two major reasons for the difference. 1) I agree with the analysis that HF bonds are generally too short for organic molecules because of the lack of dissociation consistency. However, this only applies to homolytic bond cleavage. Many organometallic bonds are dative and therefore dissociate properly. In these cases, correlation strengthens the bonds and usually shortens them as a result. 2) HF is biased toward the metal s orbital over the metal d orbital. The s orbital is more diffuse than the d orbitals and requires less correlation to describe properly. You can test this by calculating the metal s2dn-2, s1dn-1, and dn state spittings of the metal. You get much better agreement with experiment when correlation and good basis sets are used. Otherwise, there is a bias in favor of the s2dn-2 and s1dn-1 states. Because of this bias, HF descriptions of metal-ligand covalent bonds often have too much s character in them. With the addition of correlation more d character is introduced into the bond. Since the s orbital is larger than the d orbital, reducing s character and introducing more d character coincides with a shortening of the bond. Correlation may also describe charge transfer better and also add dispersion. Both could potentially lead to stronger, shorter bonds. Jason Perry First Principles Research, Inc. and Schrodinger, Inc. http://www.firstprinciples.com http://www.schrodinger.com