From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Friday, January 02, 1998 12:26 PM To: hyperchem@hyper.com Subject: Bond index calculations Happy New Year to all the readers of this newsgroup! Last year we received a number of inquiries on how to perform bond index calculations using HyperChem. This question was answered only briefly before, but will probably be added soon to the FAQ section at www.hyper.com/support . HyperChem, in its present form, does not calculate QM bond indices automatically, nor does it allow for direct access to the density matrix via Hcl-commands. The evaluation of the bond index (BI) requires the usage of the log-file information. It is exemplified here with a model-built N2 molecule, ZINDO/S, RHF singlet. Users can extend this example further to any chemical bond of interest. There are many formulas that were used to evaluate the bond index between two atoms, see A.B. Sannigrahi, Adv. Quant. Chem. 23 (1992) 302. One popular expression is that suggested by Wiberg, sum(a)sum(b) Pab*Pba, or, as originally proposed and easier to use BI = sum(a) sum(b) Pab^2 with P being the density matrix element, and the summations going over the atomic orbitals a (belonging to atom A) and b (belonging to atom B). Below is the part of the density matrix (from the HyperChem log-file) one would need in order to apply the Wiberg formula to N2, see the lower block of 4 lines: Density Matrix S N 1 Px N 1 Py N 1 Pz N 1 S N 1 1.75454 0.00000 -0.43036 0.00000 Px N 1 0.00000 1.00000 -0.00000 -0.00000 Py N 1 -0.43036 -0.00000 1.24546 -0.00000 Pz N 1 0.00000 -0.00000 -0.00000 1.00000 S N 2 0.24546 0.00000 0.43036 0.00000 Px N 2 -0.00000 1.00000 0.00000 -0.00000 Py N 2 -0.43036 0.00000 -0.75454 0.00000 Pz N 2 -0.00000 0.00000 0.00000 1.00000 BI = 0.24546 * 0.24546 + (-0.43036*-0.43036) +1.00000 * 1.00000 + 0.43036 * 0.43036 +(-0.75454 *-0.75454) + 1.00000 * 1.00000 = 3.000 (triple bond) Applying the same technique to the O2 density matrix would give BI=2 (double bond), BI=1 for H2 (single bond), etc. The Wiberg formula works also nicely for polyatomic systems with heteroatoms where non-integer bond indices usually appear, for example in HCN, the calculated C-H bond index is 0.96, and 2.94 is obtained for the triple C-N bond. The N-H bond is calculated nearly zero, as also expected. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 From: Support Sent: Tuesday, January 06, 1998 10:22 AM To: 'hyperchem@hyper.com' Subject: HELP !!! - Personal request - Subject not related From: Sorin Filip To: CHEMWEB-list , CMC-IN-HE-list , EURORESEARCH-list , EUROTALK-list , HYPERCHEM-list , NEWUSERS-list , ESRC-fellowship-schemes Subject: HELP !!! - Personal request - Subject not related Date: Tue, 06 Jan 98 10:34:18 PST (With apologies for cross-posting and for not related subject) I am a Romanian Research Assistant, chemist, 27 years old. I have Hemophilia type A, severe form and now it is impossible for me to do an ambulatory treatment because I don't have the anti-hemophilic Factor VIII. Could anyone help me with a suggestion to find a source for Factor VIII for free ? Thank you very much. Sincerely yours, Sorin. *************************************** Sorin V. FILIP Institute for Isotopic and Molecular Technology Molecular Physics Lab. Microwave Research Group 65-103 Donath St., P.O.Box 700 RO-3400, Cluj-Napoca 5 ROMANIA Tel. : +40-64-184037 / 223 Fax. : +40-64-420042 E-mail : svfilip@L30.itim-cj.ro http://www.itim-cj.ro/~micro http://www.webspawner.com/users/SorinFilip ************************************** From: Stavrev, Krassimir Sent: Tuesday, January 06, 1998 10:43 AM To: 'W R Tomlinson (in6)' Subject: RE: Bond index calculations Hi Bill, That's a good question. The Wiberg formula works fine for closed-shell singlets from orthonormal orbitals, but needs elaboration for open-shell states. One can get the bond index=2 for the *singlet* oxygen directly from the RHF or UHF density matrices using the Wiberg formula. For the true open-shell UHF triplet, however, the number of alpha and beta orbitals differ and one need to know the overlap integral between the a- and the b-sets, Sab, to calculate the bond index. In the case of non-orthogonal AOs, a formula was suggested by Giambiagi, BI= sum(a) sum(b) (PS)ab (PS)ba See A. Sannigrahi, Adv. Quant. Chem. 23 (1992) 302 I mentioned before. The paper reviews critically the various aspects of the summation over alpha- and beta- spin matrix elements. For a closed-shell system, the UHF densities are P(alpha) = P(beta) = P/2, and the summation alpha+beta is fine; however, P = P(alpha) + P(beta) is not true in general, as the alpha- and beta- orbitals are not biorthogonal and alpha- and beta- MOs differ for open-shell states. One needs to know the overlap integrals in order to evaluate the first order density D=P*S and estimate the missing terms Pab(alpha,beta) between the alpha- and the beta- AOs. Unfortunately, HyperChem does not export the overlap integrals, so one cannot use effectively the Gambiagi formula for non-singlet molecules. Singlets are presently the only possibility to examine. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: W R Tomlinson (in6) [SMTP:w.r.tomlinson@unn.ac.uk] Sent: Tuesday, January 06, 1998 5:22 AM To: Stavrev, Krassimir Subject: RE: Bond index calculations Happy New Year Dear Stavrev In regard to bond index calculations using density matrices, you state that O2 would give a bond index of 2 using the Wiberg method (and as expected from simple MO diagrams). However, for both RHF and UHF calculations my results for O2 consistently return a bond index of 1.5 (I summed the alpha and beta density matrix values for the UHF case as suggested by Joel Polowin in his message of 9 Oct, 1996) irrespective of the semiempirical Hamiltonian used. I would be grateful to know how you achieved an index of 2. The relevent part of a RHF density matrix for O2 is given below. Best regards Bill Tomlinson SO1 PxO1 PyO1 PzO1 SO2 PxO2 SO1 1.92361 -0.26535 -0.01206 0.00000 0.07639 -0.26535 PxO1 -0.26535 1.07726 -0.01922 0.00000 0.26535 -0.92068 PyO1 -0.01206 -0.01922 1.49913 0.00000 0.01206 -0.06458 PzO1 0.00000 0.00000 0.00000 1.50000 0.00000 0.00000 SO2 0.07639 0.26535 0.01206 0.00000 1.92361 0.26535 PxO2 -0.26535 -0.92068 -0.06458 0.00000 0.26535 1.07726 PyO2 -0.01206 -0.06458 0.49706 0.00000 0.01206 -0.01922 PzO2 0.00000 0.00000 0.00000 0.50000 0.00000 0.00000 BI=(0.07639)^2 + (-0.26535)^2 + (-0.01206)^2+ (0.26535)^2 + (-0.92068)^2 +(-0.06458)^2 + (0.01206)^2 + (-0.06458)^2 + (0.49706)^2 + (0.5000)^2 = 1.5 From: Maciej Turowski [turek@farmacja.amg.gda.pl] Sent: Wednesday, January 07, 1998 2:42 AM To: Sorin Filip Cc: CHEMWEB-list; CMC-IN-HE-list@ftp.hyper.com; hyperchem@hyper.com Subject: Re: Dear Filip, Is there a charity foundation that could collect money for you? If so, then please, mail their address and bank account as well as e-mail. If one had such possibility then would, for instance, help you with some money for your medicine. This could be a possibility for those not having an access to gratuite samples of the Factor VIII. Yours sincerely, Maciej Turowski Sorin Filip wrote: > , > EURORESEARCH-list , > EUROTALK-list , > HYPERCHEM-list , > NEWUSERS-list , > ESRC-fellowship-schemes > Subject: HELP !!! - Personal request - Subject not related > Date: Tue, 06 Jan 98 10:34:18 PST > Sender: listadmin@hyper.com > Precedence: bulk > Reply-To: Sorin Filip > > (With apologies for cross-posting and for not related subject) > I am a Romanian Research Assistant, chemist, 27 years old. I have > Hemophilia type A, severe form and now it is impossible for me to do an > ambulatory treatment because I don't have the anti-hemophilic Factor > VIII. > Could anyone help me with a suggestion to find a source for Factor VIII > for free ? > Thank you very much. > Sincerely yours, > Sorin. > > *************************************** > Sorin V. FILIP > Institute for Isotopic and Molecular Technology > Molecular Physics Lab. > Microwave Research Group > 65-103 Donath St., P.O.Box 700 > RO-3400, Cluj-Napoca 5 > ROMANIA > > Tel. : +40-64-184037 / 223 > Fax. : +40-64-420042 > > E-mail : svfilip@L30.itim-cj.ro > http://www.itim-cj.ro/~micro > http://www.webspawner.com/users/SorinFilip > > ************************************** From: Maciej Turowski [turek@farmacja.amg.gda.pl] Sent: Wednesday, January 07, 1998 4:20 AM To: hyperchem@hyper.com Subject: Upgrade to Hyper 5.02 Dear Users, Where, in which directory (Program, Rufiles, other) place the file UPGRADE.EXE to upgrade HyperChem to 5.02? There is a question: OK to patch C:\Hyper5? I'm a little bit confused, and grateful for your help. Sincerely, Maciej Turowski From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Wednesday, January 07, 1998 9:10 AM To: 'hyperchem@hyper.com' Subject: RE: Upgrade to Hyper 5.02 Maciej: You can place the upgrade.exe file anywhere on your hard disk and then run it. The patching program checks initially for the location of the installation directory of HyperChem, e.g. C:\Hyper5. You need to point to a file in this directory, for example, readme.txt, and then press 'Open' to proceed. To use this Maintenance Update you must have a proper installation of 5.01 or reinstall 5.01 if you have made any changes to any of its binaries, see http://www.hyper.com/news/press/maintainance.html. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: Maciej Turowski [SMTP:turek@farmacja.amg.gda.pl] Sent: Wednesday, January 07, 1998 4:20 AM To: hyperchem@hyper.com Subject: Upgrade to Hyper 5.02 Dear Users, Where, in which directory (Program, Rufiles, other) place the file UPGRADE.EXE to upgrade HyperChem to 5.02? There is a question: OK to patch C:\Hyper5? I'm a little bit confused, and grateful for your help. Sincerely, Maciej Turowski From: Brogle, Kevin [Kevin.Brogle@pharma.com] Sent: Friday, January 09, 1998 2:54 PM To: 'hyperchem@hyper.com' Subject: memory When I try to either 'print' or 'copy image' a structure with a mapped electrostatic potential surface, Hyperchem gives a low memory error after taking up 300 MB of virtual memory which it then does not release unless Hyperchem is shutdown. (These are small molecules and the printing is never accomplished.) Has anyone had experience with this problem? I am running Hyperchem 5.02 on an NT 4.0 service pack 3 workstation with 64 MB RAM. Thank you, Kevin Brogle' brogle@pharma.com From: Stavrev, Krassimir Sent: Friday, January 09, 1998 5:12 PM To: 'Brogle, Kevin' Subject: RE: memory Kevin: Is the feature you see method dependent - what method did you use, semiempirical (which one?) or ab-initio? Please send me a hin-file to look at as an example. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: Brogle, Kevin [SMTP:Kevin.Brogle@pharma.com] Sent: Friday, January 09, 1998 2:54 PM To: 'hyperchem@hyper.com' Subject: memory When I try to either 'print' or 'copy image' a structure with a mapped electrostatic potential surface, Hyperchem gives a low memory error after taking up 300 MB of virtual memory which it then does not release unless Hyperchem is shutdown. (These are small molecules and the printing is never accomplished.) Has anyone had experience with this problem? I am running Hyperchem 5.02 on an NT 4.0 service pack 3 workstation with 64 MB RAM. Thank you, Kevin Brogle' brogle@pharma.com From: Brogle, Kevin [Kevin.Brogle@pharma.com] Sent: Monday, January 12, 1998 9:37 AM To: 'Stavrev, Krassimir' Subject: RE: memory Krassimer- It happens with AM1 and PM3, I have not tried any others. With the isosurface rendering set to 'wire mesh' or 'Jorgensen-Salem' it will print but all other renderings cause the same problem. Attached is a hin-file of a small molecule with which this problem occurs. Thanks, Kevin <> > -----Original Message----- > From: Stavrev, Krassimir [SMTP:stavrev@hyper.com] > Sent: Friday, January 09, 1998 5:12 PM > To: 'Brogle, Kevin' > Subject: RE: memory > > Kevin: Is the feature you see method dependent - what method did you > use, semiempirical (which one?) or ab-initio? > Please send me a hin-file to look at as an example. > > Regards, > Krassimir > --- > Krassimir Stavrev, PhD, Director of Scientific Support > Hypercube, Inc. Florida Science and Technology Park > 1115 N.W. 4th Street Gainesville, Florida 32601 > Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 > > -----Original Message----- > From: Brogle, Kevin [SMTP:Kevin.Brogle@pharma.com] > Sent: Friday, January 09, 1998 2:54 PM > To: 'hyperchem@hyper.com' > Subject: memory > > > When I try to either 'print' or 'copy image' a structure with a mapped > electrostatic potential surface, Hyperchem gives a low memory error > after taking up 300 MB of virtual memory which it then does not > release > unless Hyperchem is shutdown. (These are small molecules and the > printing is never accomplished.) Has anyone had experience with this > problem? I am running Hyperchem 5.02 on an NT 4.0 service pack 3 > workstation with 64 MB RAM. > Thank you, > Kevin Brogle' > brogle@pharma.com From: Stavrev, Krassimir Sent: Monday, January 12, 1998 10:32 AM To: 'Brogle, Kevin' Subject: RE: memory Kevin: Thanks for the information. I can partly reproduce the problem on a Windows NT machine with 64 Mb RAM - I get the message "Out of memory. Please close other applications" when I try either Copy or Print. At the same time, the problem seem not to show up under Windows 95 where both Copy Image/F9 and Print from HyperChem 5.02 work fine for me. This could well be a NT feature. We will look further into this, will let you know when a fix exists. In the meantime, if your machine is dual boot Windows 95/NT, you may try printing the 3D-mapped isosurfaces from Windows 95 as a temporary workaround. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: Brogle, Kevin [SMTP:Kevin.Brogle@pharma.com] Sent: Monday, January 12, 1998 9:37 AM To: 'Stavrev, Krassimir' Subject: RE: memory Krassimer- It happens with AM1 and PM3, I have not tried any others. With the isosurface rendering set to 'wire mesh' or 'Jorgensen-Salem' it will print but all other renderings cause the same problem. Attached is a hin-file of a small molecule with which this problem occurs. Thanks, Kevin <> > -----Original Message----- > From: Stavrev, Krassimir [SMTP:stavrev@hyper.com] > Sent: Friday, January 09, 1998 5:12 PM > To: 'Brogle, Kevin' > Subject: RE: memory > > Kevin: Is the feature you see method dependent - what method did you > use, semiempirical (which one?) or ab-initio? > Please send me a hin-file to look at as an example. > > Regards, > Krassimir > --- > Krassimir Stavrev, PhD, Director of Scientific Support > Hypercube, Inc. Florida Science and Technology Park > 1115 N.W. 4th Street Gainesville, Florida 32601 > Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 > > -----Original Message----- > From: Brogle, Kevin [SMTP:Kevin.Brogle@pharma.com] > Sent: Friday, January 09, 1998 2:54 PM > To: 'hyperchem@hyper.com' > Subject: memory > > > When I try to either 'print' or 'copy image' a structure with a mapped > electrostatic potential surface, Hyperchem gives a low memory error > after taking up 300 MB of virtual memory which it then does not > release > unless Hyperchem is shutdown. (These are small molecules and the > printing is never accomplished.) Has anyone had experience with this > problem? I am running Hyperchem 5.02 on an NT 4.0 service pack 3 > workstation with 64 MB RAM. > Thank you, > Kevin Brogle' > brogle@pharma.com << File: diol.hin >> From: "Niall W.A.Geraghty, Chemistry Departnent, University College"@hyper.com Sent: Monday, January 12, 1998 1:08 PM Galway" Subject: Problems with Excel macros To: HYPERCHEM@HYPER.COM Date: Mon, 12 Jan 1998 18:07:12 +0000 (GMT) Sender: listadmin@hyper.com Precedence: bulk Reply-To: "Niall W.A.Geraghty, Chemistry Departnent, University College, For some time I have been happily using a variety of Excel macros based on the TOR-PLOT.xlm macro that used to be available in the Hyperchem software archive. The details were: Hyperchem, v4.5; ChemPlus v1.5; Windows, v3.11; Excel v5.0a; machine, 66MHz, 486. I have recently been transferring my work to a 233MHz Pentium II running Excel 97, under Windows 95. I was delighted to find that all the macros in the Hyperchem v4.5 "Getting Started" manual performed perfectly. However this was not the case with TOR-PLOT.xlm. Hyperchem is now taking exception to the instructions coming from the macro and flags the following: . "torsion is the name I had imaginatively given to the selected torsion angle I wanted to consider - butane was used as a model. This would appear to relate to the line: . I should point out that there is (and always has been) a space between "name" and "torsion" and that this macro, as indicated above, works perfectly under the old conditions. How do I get my two friends back on talking terms.? Niall Geraghty From: s.hogg@ic.ac.uk Sent: Tuesday, January 13, 1998 1:28 PM To: hyperchem@hyper.com Subject: Periodic Box problems This relates to HyperChem 5,02 under Win95 I have a large 'square' structure 40A x 40A x 20A that I want to enclose in a periodic box. I added the box, then removed the waters, but my 'square' structure doesn't line up with the edges of the box. It sits in the middle, and the two 'long' sides are at 45 degrees to the box, so from the top it looks like a diamond. Is there any way I can change the 'angles' of the box, and not just the size (and while we're asking, is there a way to make the box fit more 'snugly' to the structure and force the joining atoms to bond to each other without having waters between the layers)? Thanks in advance, -- Simon Hogg Imperial College, London, UK From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Tuesday, January 13, 1998 3:03 PM To: 's.hogg@ic.ac.uk' Cc: 'hyperchem@hyper.com' Subject: RE: Periodic Box problems Simon: No, you have to use a rectangular box. But you can manually align the molecule in the box by selecting it and subsequently rotating it. Of course, you have to precede the rotation by R-click + L-click on the selected molecule. One easy way to make the box fitting tighter to the molecule would be to add the molecule to a predefined box, e.g. 1. Ask for Periodic Box and define the size you want. 2. Select all and then delete all molecules, so that only the periodic box remains on screen. 3. Use File/Merge to add your molecule to the window. 4. Select and then translate/rotate your molecule to align it within the box. For Periodic box, see also the reference manual, pp. 219-223. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: s.hogg@ic.ac.uk [SMTP:s.hogg@ic.ac.uk] Sent: Tuesday, January 13, 1998 1:28 PM To: hyperchem@hyper.com Subject: Periodic Box problems This relates to HyperChem 5,02 under Win95 I have a large 'square' structure 40A x 40A x 20A that I want to enclose in a periodic box. I added the box, then removed the waters, but my 'square' structure doesn't line up with the edges of the box. It sits in the middle, and the two 'long' sides are at 45 degrees to the box, so from the top it looks like a diamond. Is there any way I can change the 'angles' of the box, and not just the size (and while we're asking, is there a way to make the box fit more 'snugly' to the structure and force the joining atoms to bond to each other without having waters between the layers)? Thanks in advance, -- Simon Hogg Imperial College, London, UK From: Journal of Molecular Modeling [jmolmod@organik.uni-erlangen.de] Sent: Wednesday, January 14, 1998 6:04 AM To: hyperchem@hyper.com Subject: J.Mol.Model.: Review open for discussion OFFICE MEMO Date: 14/1/1998 J.Mol.Model.: Review open for discussion Dear colleague, we are pleased to announce the publication of the second Journal of Molecular Modeling Review entitled "Three-Dimensional Models of Immune Cell Surface Proteins and Identification of Binding Sites" by Juergen Bajorath The review is accessible free of charge at the journals WWW-sites, http://link.springer.de/journals/jmm/ and http://www.ccc.uni-erlangen.de/info/JMOLMOD/ We encourage you to take part in the Open Discussion. The following abstract gives you a taste of the subject reviewed. ________________________________________________________________________ ________ ABSTRACT The extracellular regions of many cell surface proteins of the immune system contain distinct modules that may be linked in many different ways and are often only loosely tethered to a transmembrane segment. To identify important regions for binding in these molecules, molecular models are used to select residues for mutagenesis and to map binding sites. Many immune cell surface proteins belong to protein super- families and display only limited sequence identity compared to proteins with known three-dimensional (3D) structure, often 30% or less. Therefore, their detailed 3D structures are difficult to predict, and structure-based sequence analysis and model assessment are particularly important components of the model building process. In some cases, experimentally determined structures have made it possible to assess the accuracy of the predictions, which illustrates the opportunities and shortcomings of the approach. Herein the model-based identification of binding sites is described and representative examples are discussed. Tim Clark Henryette Roth (editor) (assistant editor) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^ ^ Journal of Molecular Modeling ^ Tel: [+49](0)9131 / 85-2948 ^ ^ Editorial Office ^ ^ ^ Computer-Chemie-Centrum ^ [+49](0)9131 / 85-6567 ^ ^ Universitaet Erlangen-Nuernberg ^ ^ ^ Naegelsbachstrasse 25 ^ Fax: [+49](0)9131 / 85-6565 ^ ^ D-91052 Erlangen ^ ^ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^ ^ email: jmolmod@organik.uni-erlangen.de ^ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^ ^ http://www.ccc.uni-erlangen.de ^ http://link.springer.de ^ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^ From: Graham Wills-Johnson [RCMWJ@bunyip.ph.rmit.edu.au] Sent: Monday, January 19, 1998 8:22 AM To: "Niall W.A.Geraghty, Chemistry Departnent, University"@ftp.hyper.com College"@hyper.com Date: Mon, 19 Jan 1998 14:33:02 EST-10ESUT Subject: Re: English as she is spoke CC: hyperchem@hyper.com > > > For some time I have been happily using a variety of Excel macros based Sender: listadmin@hyper.com Precedence: bulk Reply-To: "Graham Wills-Johnson" Australian English is not English; maybe English English is not English either! We had a problem with scripting, where, more or less at random, various script-messages from the manual were ignored - the problem only went away when we changed our system to the only REAL English. It's a bit of a nuisance when we want to run HyperChem with locally-written programs, but at least all the text-messages in the manual now work! Graham Wills-Johnson Department of Applied Chemistry RMIT Melbourne AUSTRALIA > How do I get my two friends back on talking terms.? > Niall Geraghty From: Simon Hogg [s.hogg@ic.ac.uk] Sent: Saturday, January 17, 1998 9:15 AM To: hyperchem@hyper.com Subject: PM3 with Argon Ummm, am I missing something here? I have a box of Argon Atoms, and when I do a PM3 single point energy minimisation, I get; ERROR message received from node=0 Atom1 with atmoic number 18: the SLater exponent of s orbital is less than or equal to zero, please check your parameter file pm3_1.abp What is the problem here (I do feel I am forgetting something ...) (P.S. is this the correct address for the HyperChem List, or should it be hyperchem-l@hyper.com?) -- Simon Hogg, Imperial College, London, UK Tel. +44 171 589 5111 ext. 56721 Fax. +44 171 584 3194 Email: s.hogg@ic.ac.uk Glass-List: glass-list@ic.ac.uk From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Monday, January 19, 1998 8:54 AM To: 's.hogg@ic.ac.uk' Cc: 'hyperchem@hyper.com' Subject: RE: PM3 with Argon Simon: HyperChem has no PM3 parameters for Argon - pm3_1.abp contains zeros for element #18; that causes the warning message to appear. The available parameters for SE methods are given in the Reference Manual pp. 194-196. Hyperchem@hyper.com is the correct name to submit messages to our listserver that distributes messages to a number of subscribed users; support@hyper.com is another channel for submitting questions of scientific and technical nature regarding Hypercube software. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 To: hyperchem@hyper.com From: Simon Hogg Subject: PM3 with Argon Date: Sat, 17 Jan 1998 14:14:53 +0000 Ummm, am I missing something here? I have a box of Argon Atoms, and when I do a PM3 single point energy minimisation, I get; ERROR message received from node=0 Atom1 with atmoic number 18: the SLater exponent of s orbital is less than or equal to zero, please check your parameter file pm3_1.abp What is the problem here (I do feel I am forgetting something ...) (P.S. is this the correct address for the HyperChem List, or should it be hyperchem-l@hyper.com?) -- Simon Hogg, Imperial College, London, UK Tel. +44 171 589 5111 ext. 56721 Fax. +44 171 584 3194 Email: s.hogg@ic.ac.uk Glass-List: glass-list@ic.ac.uk From: orcaro@lux.levels.unisa.edu.au Sent: Wednesday, January 21, 1998 1:01 AM To: hyperchem@hyper.com Subject: Units for Orbital properties Hi. Just wondering what the "Starting value" and "Increment" field within the "Grid Control" Dialog Box are measured in. Are they both measured in the same units? Default values for molecules we have been examining are typically of the order of 1e-20 for the starting value and 1e-7 for the increment. If they are measured on the same scale, the starting value is pretty irrelevant. Thanks. * Anthony O'Dea * * orcaro@lux.levels.unisa.edu.au * * Surface and Materials Processing * * Ian Wark Research Institute * University of South Australia * From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Wednesday, January 21, 1998 10:00 AM To: "'orcaro@lux.levels.unisa.edu.au'"@ftp.hyper.com Subject: RE: Units for Orbital properties Cc: "'hyperchem@hyper.com'" Subject: RE: Units for Orbital properties Date: Wed, 21 Jan 1998 10:12:36 -0500 Sender: listadmin@hyper.com Precedence: bulk Reply-To: "Stavrev, Krassimir" Anthony: The units for orbital plots are atomic units in terms of wave function (a0^-3/2, a0 being 0.52918 Angstroms). The effect of the initial and incremental values can be readily seen in the 2D contour plots, in conjunction with the number of contour levels that can also be set independently, see the Reference manual pp.269-270. Try the H2 molecule and visualize the 2D plot of the antibonding MO that has a positive and negative part. By varying the orientation of the molecule, the number of contour levels and the initial and incremental values, you can visualize various parts of the orbital. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: orcaro@lux.levels.unisa.edu.au [SMTP:orcaro@lux.levels.unisa.edu.au] Sent: Wednesday, January 21, 1998 1:01 AM To: hyperchem@hyper.com Subject: Units for Orbital properties Hi. Just wondering what the "Starting value" and "Increment" field within the "Grid Control" Dialog Box are measured in. Are they both measured in the same units? Default values for molecules we have been examining are typically of the order of 1e-20 for the starting value and 1e-7 for the increment. If they are measured on the same scale, the starting value is pretty irrelevant. Thanks. * Anthony O'Dea * * orcaro@lux.levels.unisa.edu.au * * Surface and Materials Processing * * Ian Wark Research Institute * University of South Australia * From: Miguel Moreno Carretero [mmoreno@ujaen.es] Sent: Friday, January 23, 1998 6:14 AM To: hyperchem@hyper.com Subject: Some questions Dear hyperchemists, I would be grateful if somebody can clearly define some terms wich appear in the *.log files when a semiempirical calculation is carried out. These terms are: CORE-CORE INTERACTIONS POINT-CHG SUM The first appears together with formation energy, bond energy, etc.. = The second and third are related with the dipole moment but I need know their exact significance. Thank you in advance. I wait for some answer. --------------------------------------------- Prof. Dr. Miguel N. Moreno Carretero Dep. Qu=EDmica Inorg=E1nica y Org=E1nica Campus Universitario "Las Lagunillas" (ed. 5) Universidad de Ja=E9n 23071-JAEN Tlf. 34 53 212150 Fax. 34 53 212186 --------------------------------------------- From: Dr. Ulrich Jordis [ujordis@pop.tuwien.ac.at] Sent: Friday, January 23, 1998 11:09 AM To: cem@relay.hyper.com; hyperchem@hyper.com Subject: How to import data files from Cambridge structural Date: Fri, 23 Jan 1998 17:18:27 +0000 Sender: listadmin@hyper.com Precedence: bulk Reply-To: "Dr. Ulrich Jordis" May I kindly ask to provide me with the details of how to import data files from Cambridge structural database for modelling within HyperChem: I thought, that starting from a known structure and altering it might give a realistic modelling result. Sincerely ...................................................................... Univ.Prof.Dr. Ulrich Jordis Prof.Dr.Ulrich Jordis Institut fuer Organische Chemie Institute of Organic Chemistry Technische Universitaet Wien Vienna Univ. of Technology (VUT) Getreidemarkt 9/154 A-1060 Vienna / AUSTRIA Tel: +43 (1) 58801.5013 Cellular: +43 664 1606176 FaxOffice: +43(1)586 6931 FaxHome: +43(1)3681 1645 From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Friday, January 23, 1998 11:18 AM To: "'Dr.UlrichJordis\"'"@ftp.hyper.com Cc: 'hyperchem@hyper.com' Subject: RE: How to import data files from Cambridge structural Date: Fri, 23 Jan 1998 11:34:08 -0500 Sender: listadmin@hyper.com Precedence: bulk Reply-To: "Stavrev, Krassimir" Ulrich: If you have ChemPlus you should be able to read the Cambridge database files directly using the Crystal builder File/Open menu. Alternatively, you may convert the *.csd input into another format that can be read by HyperChem. Babel is a popular freeware that can be used to interconvert between various formats, see our FAQ#10 at www.hyper.com/support for links to Babel download sites. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: "Dr. Ulrich Jordis" To: cem@relay.hyper.com, hyperchem@hyper.com Subject: How to import data files from Cambridge structural database Date: Fri, 23 Jan 1998 17:18:27 +0000 May I kindly ask to provide me with the details of how to import data files from Cambridge structural database for modelling within HyperChem: I thought, that starting from a known structure and altering it might give a realistic modelling result. Sincerely ...................................................................... Univ.Prof.Dr. Ulrich Jordis Prof.Dr.Ulrich Jordis Institut fuer Organische Chemie Institute of Organic Chemistry Technische Universitaet Wien Vienna Univ. of Technology (VUT) Getreidemarkt 9/154 A-1060 Vienna / AUSTRIA Tel: +43 (1) 58801.5013 Cellular: +43 664 1606176 FaxOffice: +43(1)586 6931 FaxHome: +43(1)3681 1645 From: Stavrev, Krassimir Sent: Friday, January 23, 1998 1:42 PM To: 'Miguel Moreno Carretero' Subject: RE: Some questions Miguel: The core-core interaction specified in the log-file is the sum of the repulsion energies between the valence cores (script variable scf-core-energy). This core-core repulsion contributes a positive number to the total energy. POINT-CHG is the point-charge term in the electric dipole moment; summed with other terms that come from hybridization effects gives the total dipole moment under SUM with its x, y and z-components. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: Miguel Moreno Carretero [SMTP:mmoreno@ujaen.es] Sent: Friday, January 23, 1998 6:14 AM To: hyperchem@hyper.com Subject: Some questions Dear hyperchemists, I would be grateful if somebody can clearly define some terms wich appear in the *.log files when a semiempirical calculation is carried out. These terms are: CORE-CORE INTERACTIONS POINT-CHG SUM The first appears together with formation energy, bond energy, etc.. = The second and third are related with the dipole moment but I need know their exact significance. Thank you in advance. I wait for some answer. --------------------------------------------- Prof. Dr. Miguel N. Moreno Carretero Dep. Qu=EDmica Inorg=E1nica y Org=E1nica Campus Universitario "Las Lagunillas" (ed. 5) Universidad de Ja=E9n 23071-JAEN Tlf. 34 53 212150 Fax. 34 53 212186 --------------------------------------------- From: Maciej Turowski [turek@farmacja.amg.gda.pl] Sent: Monday, January 26, 1998 6:08 AM To: Miguel Moreno Carretero Cc: hyperchem@hyper.com Subject: Re: Some questions Dear Miguel, As far as I know I may help you with two terms you asked about. If I'm wrong please anybody correct me: POINT-CHG - means point charge, i.e. the excess charge on every atom on the molecule [electrons] but it is not alway reliable since the Mulliken's type of calculation is used in HyperChem. SUM - is it the SUM in Dipole Moments? If so, then it means the total dipole moment of the system calculated [Debye]. I don't know whether it comes directly from the Mulliken charges or not. CORE-CORE Interactions - it would be nice to know for me as well. Epecially the exact physical meaning of this term and what information about structure can I extract from it? If I had time to look into the manuals ... . I hope we will hear from Hyper's support very soon and hope my explanation helped a little, Sincerely, Maciej Turowski, Ph.D. Miguel Moreno Carretero wrote: > Dear hyperchemists, > I would be grateful if somebody can clearly define some terms wich > appear in the *.log files when a semiempirical calculation is carried > out. > These terms are: > CORE-CORE INTERACTIONS > POINT-CHG > SUM > > The first appears together with formation energy, bond energy, etc.. =3D > The > second and third are related with the dipole moment but I need know > their exact significance. > Thank you in advance. I wait for some answer. > > --------------------------------------------- > Prof. Dr. Miguel N. Moreno Carretero > Dep. Qu=EDmica Inorg=E1nica y Org=E1nica > Campus Universitario "Las Lagunillas" (ed. 5) > Universidad de Ja=E9n > 23071-JAEN > Tlf. 34 53 212150 Fax. 34 53 212186 > --------------------------------------------- From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Monday, January 26, 1998 8:52 AM To: 'hyperchem@hyper.com' Subject: RE: Some questions Messages are posted also on the news server, news.hyper.com. -----Original Message----- From: Stavrev, Krassimir=20 Sent: Friday, January 23, 1998 1:42 PM To: 'Miguel Moreno Carretero' Subject: RE: Some questions Miguel: The core-core interaction specified in the log-file is the sum of the repulsion energies between the valence cores (script variable scf-core-energy). This core-core repulsion contributes a positive = number to the total energy.=20 POINT-CHG is the point-charge term in the electric dipole moment; = summed with other terms that come from hybridization effects gives the total dipole moment under SUM with its x, y and z-components. Regards, Krassimir =20 --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 -----Original Message----- From: Miguel Moreno Carretero [SMTP:mmoreno@ujaen.es] Sent: Friday, January 23, 1998 6:14 AM To: hyperchem@hyper.com Subject: Some questions Dear hyperchemists, I would be grateful if somebody can clearly define some terms wich appear in the *.log files when a semiempirical calculation is carried out. These terms are: CORE-CORE INTERACTIONS POINT-CHG SUM The first appears together with formation energy, bond energy, etc.. = =3D The second and third are related with the dipole moment but I need = know their exact significance. Thank you in advance. I wait for some answer. --------------------------------------------- Prof. Dr. Miguel N. Moreno Carretero Dep. Qu=EDmica Inorg=E1nica y Org=E1nica Campus Universitario "Las Lagunillas" (ed. 5) Universidad de Ja=E9n 23071-JAEN Tlf. 34 53 212150 Fax. 34 53 212186 --------------------------------------------- From: Maciej Turowski [turek@farmacja.amg.gda.pl] Sent: Tuesday, January 27, 1998 9:48 AM To: hyperchem@hyper.com Subject: hyper 5 bugs Date: Tue, 27 Jan 1998 15:47:59 +0100 (MET) Sender: listadmin@hyper.com Precedence: bulk Reply-To: Maciej Turowski Dear users, Dear support, I definately notice version 5.0, 5.01 or 5.02 as much worse for calculations than the version 4.5 or 4.0. Below please find why. Primo: During semi-empirical calculations upon certain molecular systems (not large, max. up to 100 atoms in molecule) we obtain deformated bonds, with lenght of several Angstroms (50 for instance) or at the end (in the morning for example when structure is leaved overnight for optimization) we got NOT the entire structure but either isolated atoms or a WEB of bonds mixed together, a real mess. NEVER it happened with Hyper 4.5 (we used the same structures and methods). Secundo: Molecular dynamics for simulating reactions, for example: CH3Br + OH- -> CH3OH + Br-when simulated by MD with use of AM1 or PM3 method, in Hyper 5.02 (or 5.01) we got a collection of separated atoms instead of products (after collision of two substrats) and very rarely the result of the reaction is shown as above. In one computer MD doesn't want to give movement to CH3Br despite we have set the velocity at 100 or 200 A/ps. When doing THE SAME (exactly the same) calculations in Hyper 4.0 or 4.5 we ALWAYS got the reaction which is correct (as above) and there are NO problems at all!!! Tertio: there are scripts (scr) working well in version 4.5 but displaying errors in version 5.02. The same scripts in two versions of HyperChem and it seems that the older one works better. Quarto: Wo do not see an improvement (regarding those exemplary bugs) after the upgrade from 5.01 to 5.02. If someone wishes we may send the HIN files of those damaged structures (by version 5) and the scripts as well. Also if anybody have met similar problem please write about it to this discussion list. It is very irritating to pay for the newer version quite significant money and see that they are lost, because many calculations must be done with the older one. The machines we use are quite decent (Pentium 166, 32 MB RAM) and used the same for both versions. One I'd like to notice here - the graphics in Hyper 5 is really wonderfull. Much better than previously. But ... what gen I get from this excellent graphics when I couldn't get proper results??? Hoping to hear from somebody, Yours sincerely, Maciej Turowski, Ph.D. Dept. Biopharmaceutics & Pharmacodynamics Medical University of Gdansk Poland From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Tuesday, January 27, 1998 10:49 AM To: 'turek@farmacja.amg.gda.pl' Cc: 'hyperchem@hyper.com' Subject: RE: hyper 5 bugs Maciej: You can send us the hin-files and scripts you mentioned to support@hyper.com. Please provide also a brief description of the computational setup. The list of bug fixes in HyperChem 5.02 is posted on the Web at http://www.hyper.com/support/software/updates/hyperchem5/501to502.htm. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 Date: Tue, 27 Jan 1998 15:47:59 +0100 (MET) From: Maciej Turowski To: hyperchem@hyper.com Subject: hyper 5 bugs Message-ID: Date: Tue, 27 Jan 1998 15:47:59 +0100 (MET) Dear users, Dear support, I definately notice version 5.0, 5.01 or 5.02 as much worse for calculations than the version 4.5 or 4.0. Below please find why. Primo: During semi-empirical calculations upon certain molecular systems (not large, max. up to 100 atoms in molecule) we obtain deformated bonds, with lenght of several Angstroms (50 for instance) or at the end (in the morning for example when structure is leaved overnight for optimization) we got NOT the entire structure but either isolated atoms or a WEB of bonds mixed together, a real mess. NEVER it happened with Hyper 4.5 (we used the same structures and methods). Secundo: Molecular dynamics for simulating reactions, for example: CH3Br + OH- -> CH3OH + Br-when simulated by MD with use of AM1 or PM3 method, in Hyper 5.02 (or 5.01) we got a collection of separated atoms instead of products (after collision of two substrats) and very rarely the result of the reaction is shown as above. In one computer MD doesn't want to give movement to CH3Br despite we have set the velocity at 100 or 200 A/ps. When doing THE SAME (exactly the same) calculations in Hyper 4.0 or 4.5 we ALWAYS got the reaction which is correct (as above) and there are NO problems at all!!! Tertio: there are scripts (scr) working well in version 4.5 but displaying errors in version 5.02. The same scripts in two versions of HyperChem and it seems that the older one works better. Quarto: Wo do not see an improvement (regarding those exemplary bugs) after the upgrade from 5.01 to 5.02. If someone wishes we may send the HIN files of those damaged structures (by version 5) and the scripts as well. Also if anybody have met similar problem please write about it to this discussion list. It is very irritating to pay for the newer version quite significant money and see that they are lost, because many calculations must be done with the older one. The machines we use are quite decent (Pentium 166, 32 MB RAM) and used the same for both versions. One I'd like to notice here - the graphics in Hyper 5 is really wonderfull. Much better than previously. But ... what gen I get from this excellent graphics when I couldn't get proper results??? Hoping to hear from somebody, Yours sincerely, Maciej Turowski, Ph.D. Dept. Biopharmaceutics & Pharmacodynamics Medical University of Gdansk Poland From: listadmin@hyper.com Sent: Wednesday, January 28, 1998 5:11 AM To: hyperchem@hyper.com Subject: Symmetry To: hyperchem@hyper.com Subject: Symmetry Date: Wed, 28 Jan 1998 11:11:00 +0100 Sender: listadmin@hyper.com Precedence: bulk Reply-To: Di Vona X-Mailer: Mozilla 4.03 [it] (Win95; I) MIME-Version: 1.0 Rome, january28th,1998 Dear Hyperchemists, I would like to run a semiempirical ZINDO/1 calculation on metal derivatives of cyclic tetrapyrroles. The symmetry of the systems is usually not higher than Cs and I would like to impose a C2v or higher symmetry. Does anyone know if it is feasible? And how? Thank you in advance for your cooperation. Ciao Luisa **************************************** Dr Maria Luisa Di Vona Dipartimento di Scienze e Tecnologie Chimiche Universit=E0 di Roma Tor Vergata 00133 Rome, Italy Phone: 39.6.72594382 Fax: 39.6.72594328 e-mail: divona@uniroma2.it ***************************************** From: Stavrev, Krassimir [stavrev@hyper.com] Sent: Wednesday, January 28, 1998 10:56 AM To: 'divona@uniroma2.it' Cc: 'hyperchem@hyper.com' Subject: RE: Symmetry Luisa: HyperChem does not control/impose symmetry directly. The constraint options in the build menu can help you construct a symmetrical structure. However, when you start a geometry optimization (using ZINDO/1 or other methods), the structure may, or may not, remain symmetrical as built. If it departs significantly from the one you wish to examine, you can still use Restraints, so as to keep certain bonding elements frozen and have the symmetry preserved to a certain extent. Let me know if you need an example on how to use Restraints. Regards, Krassimir --- Krassimir Stavrev, PhD, Director of Scientific Support Hypercube, Inc. Florida Science and Technology Park 1115 N.W. 4th Street Gainesville, Florida 32601 Voice:(352) 371-7744/(800) 960-1871 Fax:(352) 371-3662 From: Di Vona X-Mailer: Mozilla 4.03 [it] (Win95; I) MIME-Version: 1.0 To: hyperchem@hyper.com Subject: Symmetry Date: Wed, 28 Jan 1998 11:11:00 +0100 Rome, january28th,1998 Dear Hyperchemists, I would like to run a semiempirical ZINDO/1 calculation on metal derivatives of cyclic tetrapyrroles. The symmetry of the systems is usually not higher than Cs and I would like to impose a C2v or higher symmetry. Does anyone know if it is feasible? And how? Thank you in advance for your cooperation. Ciao Luisa **************************************** Dr Maria Luisa Di Vona Dipartimento di Scienze e Tecnologie Chimiche Università di Roma Tor Vergata 00133 Rome, Italy Phone: 39.6.72594382 Fax: 39.6.72594328 e-mail: divona@uniroma2.it ***************************************** From: Joe Rizzolo [rizzolo@norwich.edu] Sent: Wednesday, January 28, 1998 5:11 PM To: Di Vona Cc: hyperchem@hyper.com Subject: Re: Symmetry Hi Prof. Di Vona !!! Regarding symmetry of metal tetrapyroles, I have done similar calculations on metal porphyrins using ZINDO/1. I have found that the key is to ensure the proper conjugation within the free base or diannion ring. If I remember correctly, after double clicking and in some cases deleting some bonds, I then performed a MM+ calculation first. Once this is done, the metal can then be added and for the first transition series, all of the porphyrins that I have done, did indeed arrive at the D4h symmetry, although it is not a minimum but a transition state structure. While I am at it I would also include the comment that apparently some M.O. model chemistries do not result in the proper alternation of bonds and partial charges within the porphyrin free base. This I know is true for the AM1. From what I understand, this is an example of the "singlet instability problem" (see for example, Reynolds, Charles H., J. Org. Chem., 1988, 53 6061-6064). Apparently this problem is due electron correlation which is necessary when dealing with conjugated systems. Interestingly, the proper symmetry and bond length/partial charge separation was achieved when using a UHF model. I hope this helps. Thanks for the opportunity to try to help. Take care everybody........ Joe Rizzolo Norwich University From: listadmin@hyper.com Sent: Friday, January 30, 1998 8:34 AM To: Stavrev, Krassimir Cc: Mirek Sopek; hyperchem@hyper.com Subject: Re: hyper 5 bugs oscar@bilbo.edu.uy, cyclolab@compuserve.com Subject: Re: hyper 5 bugs Date: Fri, 30 Jan 1998 12:52:30 +0100 Sender: listadmin@hyper.com Precedence: bulk Reply-To: Support >Dear Krassimir, Dear Mirek, Dear Hyperchemists, Your script, corrected by yourselves is working OK until ... .But we have to start from the beginning: You are writing that: "There were several trivial mistakes that were introduced to the scripts, probably by the text editor that was used, such as broken comment lines and uncorrected paths in the hin-files". If so, why it does work in Hyper 4.5 without problems? But, nevertheless, I ran your script (mac.scr) and everything went well until the moment it did exactly the same as previously. So it seems this mistake continues. The problem regards docking to the nucleic acid structure. Resulting file (from MD from the script mac.scr) is attached as DZIWNE.HIN. About semi-empirical calculations - the structure enclosed is not unique. Important: this WEB of bonds (after optimization) does NOT happen always. Only sometimes. It is more strange therefore - once the optimization runs O.K., second time the resulting structures were destroyed totally. Only in Hyper 5 and above this situation happened. The procedure of calculations were as follows: 1) After running Model Builder the structure was "eqiupped" with atomic charges (either from ChemPlus or Single Point AM1), 2) The geometry optimization (MM+ force field) was then performed, 3) Semi empirical optimization followed with following settings: Total charge = 0 Spin pairing = RHF State = Lowest Convergence limit = 0.01 Accelerate convergence = YES Algorithm = Polak-Ribiere RMS gradient = 0.1 kcal/mol In vaccuo = YES I am enclosing the molecule before and after optimization again (nhexylbenzene.hin and nhexylbenzene-after.hin). Perhaps it is a question of the hardware but I don't think so since HyperChem 5.02 is running on more than one machine. But ... . And finally, the reaction simulation: OH- + CH3Br >>>> CH3OH + Br- The file JUST BEFORE running MD is attached as REAKCJA.HIN The snapshot file is REAKCJA.snp. Now it looks O.K. but believe me, at the same conditions I received all structures destroyed (i.e. atoms separated) and collision was so strong once that the oxygen atom DROVE INto the carbon atom!!! It looked almost like nuclear reaction! Try this, perhaps someone will get similar strange results. The thing is that sometimes it repeats sometimes not (i.e. is correct). I have set MD options as follows: 1) The molecules were builded by model builder only and translated in space to fit the reaction path (as saved in file reakcja.hin), 2) The system was set to semi-empirical PM3, total charge -1, multiplicity 1, RHF, convergence limit 0.1. 3) Then the oxygen atom (in OH) was selected as POINT. 4) Then the CH3Br molecule has received velocity (SET VELOCITY - AT POINT) at 100 A/ps. 5) Then MD was performed: Heat time = 0 Run time = 1 Cool time = 0 Step size = 0.001 ps Starting temp. = 0 Simulation temp. = 0 Final temp. = 0 In vacuo Constant temp. = NO Restart = NO Happy calculations!!! Looking forward for the messages, Sincerely, Maciej From: Support Sent: Friday, January 30, 1998 9:17 AM To: 'Maciej Turowski' Subject: RE: hyper 5 bugs Your posting to the HyperChem newsgroup contains large supplementary material and cannot be distributed with it. We will post only the text message. The full submission is placed on the Hypercube NNTP server, news.hyper.com. -----Original Message----- From: Maciej Turowski [SMTP:turek@farmacja.amg.gda.pl] Sent: Friday, January 30, 1998 6:53 AM To: Stavrev, Krassimir Cc: Mirek Sopek; hyperchem@hyper.com; oscar@bilbo.edu.uy; cyclolab@compuserve.com Subject: Re: hyper 5 bugs > Dear Krassimir, Dear Mirek, Dear Hyperchemists, Your script, corrected by yourselves is working OK until ... .But we have to start from the beginning: You are writing that: "There were several trivial mistakes that were introduced to the scripts, probably by the text editor that was used, such as broken comment lines and uncorrected paths in the hin-files". If so, why it does work in Hyper 4.5 without problems? But, nevertheless, I ran your script (mac.scr) and everything went well until the moment it did exactly the same as previously. So it seems this mistake continues. The problem regards docking to the nucleic acid structure. Resulting file (from MD from the script mac.scr) is attached as DZIWNE.HIN. About semi-empirical calculations - the structure enclosed is not unique. Important: this WEB of bonds (after optimization) does NOT happen always. Only sometimes. It is more strange therefore - once the optimization runs O.K., second time the resulting structures were destroyed totally. Only in Hyper 5 and above this situation happened. The procedure of calculations were as follows: 1) After running Model Builder the structure was "eqiupped" with atomic charges (either from ChemPlus or Single Point AM1), 2) The geometry optimization (MM+ force field) was then performed, 3) Semi empirical optimization followed with following settings: Total charge = 0 Spin pairing = RHF State = Lowest Convergence limit = 0.01 Accelerate convergence = YES Algorithm = Polak-Ribiere RMS gradient = 0.1 kcal/mol In vaccuo = YES I am enclosing the molecule before and after optimization again (nhexylbenzene.hin and nhexylbenzene-after.hin). Perhaps it is a question of the hardware but I don't think so since HyperChem 5.02 is running on more than one machine. But ... . And finally, the reaction simulation: OH- + CH3Br >>>> CH3OH + Br- The file JUST BEFORE running MD is attached as REAKCJA.HIN The snapshot file is REAKCJA.snp. Now it looks O.K. but believe me, at the same conditions I received all structures destroyed (i.e. atoms separated) and collision was so strong once that the oxygen atom DROVE INto the carbon atom!!! It looked almost like nuclear reaction! Try this, perhaps someone will get similar strange results. The thing is that sometimes it repeats sometimes not (i.e. is correct). I have set MD options as follows: 1) The molecules were builded by model builder only and translated in space to fit the reaction path (as saved in file reakcja.hin), 2) The system was set to semi-empirical PM3, total charge -1, multiplicity 1, RHF, convergence limit 0.1. 3) Then the oxygen atom (in OH) was selected as POINT. 4) Then the CH3Br molecule has received velocity (SET VELOCITY - AT POINT) at 100 A/ps. 5) Then MD was performed: Heat time = 0 Run time = 1 Cool time = 0 Step size = 0.001 ps Starting temp. = 0 Simulation temp. = 0 Final temp. = 0 In vacuo Constant temp. = NO Restart = NO Happy calculations!!! Looking forward for the messages, Sincerely, Maciej << File: nhexylbenzene.hin >> << File: nhexyllbenzen-after.hin >> << File: reakcja.hin >> << File: reakcja.snp >> << File: dziwne.hin >> << File: mac.scr >>