From uunet.ca!ciens.ula.ve!alejo@hyper Mon Apr 4 20:54:57 1994 Date: Mon, 4 Apr 1994 14:15:15 -0400 From: ALEJANDRO MUJICA To: hyperchem@HYPER.COM Subject: homology with hyperchem? Cc: alejo@ciens.ula.ve Hello everybody, Is there any expierience among you using Hyperchem R3 in homology problems? We are using Hyperchem in basic protein modeling and we would like to go further... Any information may be useful. Thanks in advance. Alejandro #################################################### Alejandro Mujica e-mail: alejo@ula.ve Departamento de Biologia Facultad de Ciencias Universidad de Los Andes Merida - Venezuela Fax: (58 74) 401379 ##################################################### ________________________________________________________________________ From uunet.ca!ucdavis.edu!dwgruenwedel@hyper Wed Apr 6 21:21:19 1994 From: ucdavis.edu!dwgruenwedel@hyper.uucp Date: Wed, 6 Apr 1994 19:17:14 -0400 To: hyperchem@HYPER.COM Subject: Computations Involving Chemical Elements above 54 (Xe) I am interested in performing semi-empirical calculations on the mercury (Hg(II), atomic number 80) complexes of purines and pyrimidines. Since Hg(II) has been subjected to parametrization (e.g., in PM3), I would have thought that by incorporating the requisite parameters of Hg(II) into the Hyperchem PM3_*.ABG files one should be able to perform the desired computations. Thus far, I have received only error messages when attempting to execute them. Questions: (1) Why does one get the error messages? In view of Hyperchem's 'open architecture', one should be able to edit and modify the files in question. (2) Does anyone have an answer to the problem(s)? (3) Is anyone aware of software that would permit semi-empirical computations on third-row transition metals (including Hg)? Thank you for your help. ************************************************************* * Prof. Dieter W. Gruenwedel (916) 752-2155 (voice) * * Food Science & Technology (916) 752-4759 (FAX) * * 109 Food Science Bldg. dwgruenwedel@ucdavis.edu * * Univ. of California, Davis, CA 95616, USA * ************************************************************* ________________________________________________________________________ From uunet.ca!UCDAVIS.EDU!dwgruenwedel@hyper Thu Apr 7 13:08:01 1994 From: UCDAVIS.EDU!dwgruenwedel@hyper.uucp Date: Thu, 7 Apr 1994 12:21:39 -0400 To: hyperchem@HYPER.COM Subject: There seems to have been a "delivery problem". Hence, I repeat everything. ************************************************************* * Prof. Dieter W. Gruenwedel (916) 752-2155 (voice) * * Food Science & Technology (916) 752-4759 (FAX) * * 109 Food Science Bldg. dwgruenwedel@ucdavis.edu * * Univ. of California, Davis, CA 95616, USA * ************************************************************* ------------------------- Forwarded Message Follows ------------------------- To: hyperchem@hyper.com Subject: Computations Involving Chemical Elements above 54 (Xe) Cc: I am interested in performing semi-empirical calculations on the mercury (Hg(II), atomic number 80) complexes of purines and pyrimidines. Since Hg(II) has been subjected to parametrization (e.g., in PM3), I would have thought that by incorporating the requisite parameters of Hg(II) into the Hyperchem PM3_*.ABG files one should be able to perform the desired computations. Thus far, I have received only error messages when attempting to execute them. Questions: (1) Why does one get the error messages? In view of Hyperchem's 'open architecture', one should be able to edit and modify the files in question. (2) Does anyone have an answer to the problem(s)? (3) Is anyone aware of software that would permit semi-empirical computations on third-row transition metals (including Hg)? Thank you for your help. ************************************************************* * Prof. Dieter W. Gruenwedel (916) 752-2155 (voice) * * Food Science & Technology (916) 752-4759 (FAX) * * 109 Food Science Bldg. dwgruenwedel@ucdavis.edu * * Univ. of California, Davis, CA 95616, USA * ************************************************************* ________________________________________________________________________ From slee@hyper Thu Apr 7 13:24:26 1994 From: slee@hyper.uucp (Thomas Slee) Subject: Re: To: dwgruenwedel@UCDAVIS.EDU Date: Thu, 7 Apr 1994 13:13:48 -0400 Cc: hyper!hyperchem@uunet.ca Dieter W. Gruenwedel writes.... | I am interested in performing semi-empirical calculations on the | mercury (Hg(II), atomic number 80) complexes of purines and | pyrimidines. Since Hg(II) has been subjected to parametrization | (e.g., in PM3), I would have thought that by incorporating the | requisite parameters of Hg(II) into the Hyperchem PM3_*.ABG files | one should be able to perform the desired computations. Thus far, | I have received only error messages when attempting to execute | them. | | Questions: | | (1) Why does one get the error messages? In view of Hyperchem's | 'open architecture', one should be able to edit and modify the | files in question. | | (2) Does anyone have an answer to the problem(s)? | | (3) Is anyone aware of software that would permit semi-empirical | computations on third-row transition metals (including Hg)? | | Thank you for your help. HyperChem is currently able to tackle elements up to Xenon. Beyond that, the problem is not with parameters, but with some of the integrals required to do the calculations. This is why adding parameters does not help. As for software that does do Hg: MOPAC will do this calculation for you using PM3 which, as you say, is one of the few heavier elements to be parameterised for accurate semi-empirical methods. There may be other commercial software, but I will not comment on that. Yours Tom Slee -- Tom Slee, Hypercube Inc., 419 Phillip St., Waterloo, Ont., Canada, N2L 3X2 E-Mail: slee@hyper.com, Tel. (519)725-4040, Fax. (519)725-5193. ________________________________________________________________________ From uunet.ca!biochemistry.BIOC.CWRU.Edu!anderson@hyper Thu Apr 14 10:23:00 1994 From: "Vernon Anderson" Date: Thu, 14 Apr 1994 10:04:02 -0400 To: hyperchem@hyper.com Subject: Hyperchem under Windows NT on DEC alpha I have a colleague who said he saw Hyperchem running under NT on a DEC alpha at the ACS meeting in San Diego. I was considering the purchase of a DEC alpha system and asked the sal;esman if they could run Windows NT, and the answer was , "no", except for the 2000-500 NT server system. I would like to know if it is possible to run WIndows NT on a DEC alpha platform and if Hyperchem version 3 for Windows will then run under it, or if I will have to purchase a new version of Hyperchem. Thanks. -Vernon- 10900 Euclid Avenue 216-368-2599, fax 216-368-4544 Case Western Reserve University anderson@biochemistry.cwru.edu Cleveland, OH 44106-4935 ________________________________________________________________________ From slee@hyper Thu Apr 14 11:18:19 1994 From: slee@hyper.uucp (Thomas Slee) Subject: Re: Hyperchem under Windows NT on DEC alpha To: hyperchem@hyper.com Date: Thu, 14 Apr 1994 11:16:45 -0400 Vernon Anderson writes... | | I have a colleague who said he saw Hyperchem running under NT on a DEC | alpha at the ACS meeting in San Diego. I was considering the purchase of a | DEC alpha system and asked the sal;esman if they could run Windows NT, and | the answer was , "no", except for the 2000-500 NT server system. I would | like to know if it is possible to run WIndows NT on a DEC alpha platform | and if Hyperchem version 3 for Windows will then run under it, or if I will | have to purchase a new version of Hyperchem. | Thanks. | We did demonstrate a pre-production version of HyperChem for NT on the DEC Alpha workstation at the ACS meeting. I am puzzled by the Dec (?) salesman who said they could not run Windows NT: you do have to choose whether to install NT or Dec's OSF Unix on the workstation, and you cannot run both of them simultaneously, but you can certainly buy a Dec alpha workstation running NT: they are advertising it extensively. No version of HyperChem for Windows will run under NT on the Alpha platform. HyperChem requires Windows running in enhanced mode because it contains 32-bit programs, and NT on RISC chips does not support enhance mode. We are far enough along in developing a version of HyperChem specifically for NT that we demonstrated it at the San Diego show, but we are not yet far enough along to announce a release date. Hope this clear up some of your questions. Yours Tom Slee -- Tom Slee, Hypercube Inc., 419 Phillip St., Waterloo, Ont., Canada, N2L 3X2 E-Mail: slee@hyper.com, Tel. (519)725-4040, Fax. (519)725-5193. ________________________________________________________________________ From uunet.ca!surrey.ac.uk!chp1aa@hyper Thu Apr 14 12:07:04 1994 From: Mr Andrew D Allen Subject: Batch jobs on SG version To: hyperchem@HYPER.COM Date: Thu, 14 Apr 1994 11:50:44 -0400 On the SG version (ver2.0) of hyperchem is it posible to start a job as a background or batch process? Andy. -- ################################################################################ Structural and Computation Chemistry Group__________chp1aa@uk.ac.surrey - JANET. Department of Chemistry_____________________________phone__+44(483)-300800-2632. University of Surrey________________________________fax_________+44(483)-300803. Guildford, Surrey, GU2 5XH, UK______________________ftp____________131.227.110.2 ################################################################################ ________________________________________________________________________ From uunet.ca!mitr.p.lod.edu.pl!plonkaw@hyper Mon Apr 18 05:13:24 1994 From: mitr.p.lod.edu.pl!plonkaw@hyper.uucp (Wojciech Plonka home (48 42) 846780) To: hyperchem@hyper.com Subject: 3mol Date: Mon, 18 Apr 1994 04:53:43 -0400 Dear netters, I would like to answer all of you who were interested in commersial version of stereocscopic viewer for HIN files: Due to changeover of the HyperChem bussines from Autodesk to Hypercube, my situation, as the Registered Developer has changed too. 3mol surely will be made commercial, but today I can't say anything about the availability date and conditions. I appologise for the inconvienience, hope the program will get better when released. Comments welcome. Demo available on host ftp.funet.fi /pub/sci/molbio/msdos/3moldemo.zip Wojciech Plonka plonkaw@mitr.p.lod.edu.pl ________________________________________________________________________ From uunet.ca!KENTVM.KENT.EDU!IMCVEY@hyper Wed Apr 20 15:12:07 1994 Date: Wed, 20 Apr 1994 14:23:27 -0400 From: Iain McVey Subject: Lop P values make no sense To: hyperchem@hyper.com We recently purchased chem plus especially for its ability to do QSAR calculati ons. BUT when we do Log P calculations we are getting some very odd answers. For example according to QSAR log P for Acetate ions (CH3COO-) is more negative than Log P for acetic acid (CH3COOH). Is there a problem with QSAR or am I missing something crucial. Also I just received advertising literature about a new improved log P calculator, does this new product address this problem, or am I sunk? Would adding counter ions to the ionic systems fix the problem? ________________________________________________________________________ From slee@hyper Thu Apr 21 12:34:53 1994 From: slee@hyper.uucp (Thomas Slee) To: hyper!hyperchem@uunet.ca Date: Thu, 21 Apr 1994 12:33:18 -0400 Iain McVey writes... | | When we do Log P calculations [with ChemPlus] we are getting some | very odd answers. | | For example according to QSAR Properties, log P for Acetate ions (CH3COO-) is | more negative than Log P for acetic acid (CH3COOH). Is there a problem with | QSAR Properties or am I missing something crucial. I have three comments about this observation: 1). The value for acetic acid produced by ChemPlus (-0.17) is actually pretty good, with the experimental value being around -0.17 or -0.31 (two different experimental values taken from Leo's paper back in 1971 (A. Leo et al., Partition Coefficients and their Uses, Chem. Revs. 1971, 71, 525)). 2). It is not clear to me that the Log P for an ion has a meaning, by itself, but I'm quite willing to be corrected on this. Certainly I haven't seen ions included in the standard tabulations of values, or referred to in the various methods for estimating Log P that I have seen. 3). ChemPlus's method for calculating Log P is an atom based fragment method as developed and published by Ghose and Crippen. It assigns types to different atoms based on their connectivities and neighbours. It does not attempt to cover the cases of ions. I'm not aware of other published methods that try to cover ionic species, so I strongly suspect that this limitation, if it is one (see (2)), is common to most programs that predict Log P values. | | Would adding counter ions to the ionic systems fix the problem? | It would not. The value for the hydrogen ion would be arbitrary, for the same reasons as in (3). | | Also I just received advertising literature about a new improved log P | calculator, does this new product address this problem, or am I sunk? | The product I believe you are referring to is not a Hypercube product. Perhaps the developers of this product may post something here to clarify. I hope these comments are helpful. Please do correct me if I'm wrong about the ionic values point. Yours Tom Slee -- Tom Slee, Hypercube Inc., 419 Phillip St., Waterloo, Ont., Canada, N2L 3X2 E-Mail: slee@hyper.com, Tel. (519)725-4040, Fax. (519)725-5193. ________________________________________________________________________ From uunet.ca!VTVM1.CC.VT.EDU!VTVM1.CC.VT.EDU!RAMSPEC@hyper Thu Apr 21 12:38:54 1994 Date: Thu, 21 Apr 1994 11:06:26 -0400 From: Alan Schick Subject: Vibrations of symmetric molecs. To: hyperchem users group Hyperchem users, I am interested in the effects of asymmetric peripheral substitution on the vibrational and electronic properties of porphyrin molecules. To this end I am trying to use Hyperchem's AM1 (or PM3) method to compute the normal modes. I have Release 3 (no ChemPlus). As one would expect, even for the smallest possible porphyrin (no peripheral substituents) the calculations, including geometry minimization (to gradient < 0.01) and subsequent vibrational calculation, take VERY LONG times; so rather than spend literally weeks at "wheel spinning", I thought I might seek advice early on. I have noticed some fundamental problems/observations with my results. Any comments/advice would be greatly appreciated: Symmetry problems: I know that Hyperchem does not rigorously deal with symmetry, and perhaps as a result, I am having symmetry problems. The free-base porphin (unsubstituted porphyrin) should have D2h symmetry. Using the model builder, a very reasonable-looking structure is generated when I define (i) two opposite pyrrole nitrogens to contain the central protons, and (ii) the conjugation pathway (dotted double bonds) includes the Cb-Cb, Cb-Ca bonds of these same pyrroles, all Ca-Cm bonds and the Ca-N bonds of the other two pyrroles. The Cb-Cb bonds of the other two pyrroles are defined as double bonds, not in the conjugation pathway. This structure, if it is planar, is rigorously D2h. The model builder, though, does not generate a rigorously D2h structure, with equivalent bonds on opposite sides of the structure being off by as much as ~10%, and some equivalent bond angles ranging by 1-2 degrees. The structure is, however, planar. The geometry optimization does not rectify this lack of symmetry, and in fact slightly worsens it. The result is that the normal modes are, of course, not recognizable with respect to the many normal modes calcs available from the literature (using standard FG matrix methods). And the calc modes are nowhere near symmetric. Questions: Is there anything to be done for this? I can explicitly generate a .HIN file that starts off symmetric, but will the geometry optimization take that symmetry away? Is there a better method than AM1 or PM3 to be used for this purpose (given that even these methods often give finger-print frequencies 100 to 200 cm-1 off anyway)? Another concern: The pyrrole nitrogens that contain the central protons are not really part of the conjugation path, and as I recall, tautomerism in the porphyrin core is slow enough that the core nitrogens are not equivalent -- so the proton-associated nitrogens should show at least some sp3 hybrid-like character -- i.e., they would not explicitly be planar. Yet the model builder and AM1 both optimize the geometry to be completely planar. PM3 does, in fact, "pucker" the proton of one pyrrole out of plane, BUT ONLY ONE! Thus, symmetry is lost again. I am presuming that all of my problems stem from the fact that Hyperchem does not explicitly deal with symmetry elements during calculations, so maybe my bottom-line question is: What can I do about it? Does either ChemPlus or the new upcomming release deal with symmetry any better? I can send .HIN files to anyone interested in having a look. Thanks for any assistance I can get. Alan ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ G. Alan Schick ,/\ ramspec@vtvm1.cc.vt.edu Assistant Professor ,/ `\/`'\ Voice: 703-231-8215 Dept. of Chemistry ,/\/ / Fax: 703-231-3255 Virginia Tech ,/ <_ ._. Blacksburg, VA 24061-0212 / `\ / / ,/\_ _/' `-`-\ |/ ,/' `_'_' * Blacksburg \_> ' ,/' \__>_/\ +----------------------------------------------+ ________________________________________________________________________ From slee@hyper Thu Apr 21 15:21:08 1994 From: slee@hyper.uucp (Thomas Slee) Subject: HyperChem Release 4 Now Available To: hyper!hyperchem@uunet.ca Date: Thu, 21 Apr 1994 15:17:44 -0400 HyperChem Release 4 for Windows Now Available ============================================= Hypercube is pleased to announce that HyperChem Release 4 for Windows shipped on April 18, and is available from your HyperChem reseller. This release marks the return to Hypercube from Autodesk of the rights to sell and market HyperChem, and includes the following improvements: o HyperChem's interface now has a 3D look, and the tools have a convenient pop-up description. o The calculations have been speeded up, with some semi-empirical calculations a factor of two faster than Release 3. o HyperChem is now available in a Network version, suitable for running on local area networks. o The documentation has been updated to include all material new since Release 2 into the proper books. o Several bugs from Release 3 have been fixed. Suggested retail prices are, in US dollars: Purchase: $1995 ($995 Government and Educational Institutions) Upgrade from Release 3: $195 Upgrade from Release 2: $495 Network pricing is available on request. For more details concerning this release, please send e-mail to info@hyper.com. ________________________________________________________________________ From uunet.ca!phri.nyu.edu!liu@hyper Mon Apr 25 19:10:50 1994 Date: Mon, 25 Apr 1994 18:40:13 -0400 From: phri.nyu.edu!liu@hyper.uucp (David J. Liu) To: hyperchem@hyper.com I was offered a chance to test-drive HyperChem 3.0 for Windows, and here are my comments, or, my 'wish list'. After using various features of this software, I have found that the user-interface needs much refinement to be satisfactory, below is a partial list in random order. 1. there is not an option of type-in command with history recall, I expected the log file be easily converted to script, but it did not contain a list of operations executed; 2. it is difficult to select one molecule among several loaded and save it in a new file; 3. moving molecules in 3-D space with consideration of inter-molecule interactions is not convenient; 4. rendering of selected objects only is not implemented, i.e., every atom has to be of the same style, all sphere, all stick, all dot, etc.; 5. real-time adjustment of bond-length, bond-angle, etc., with instant display of the relevant parameter(s) is not available. In addition, ChemPlus copies files to SYSTEM subdirectory without providing a list of such files, although it is not difficult to recognize them. Despite these, however, no bug has been found, the problem is generally with the ideas of software designing. The features in the above "wish list" are important, in my view, for building a starting model prior to refinement, and their implementation is easy, judged from my experience of programming. Sincerely, David J. Liu ________________________________________________________________________ From uunet.ca!ALIJKU11.EDVZ.Uni-Linz.AC.AT!jk.uni-linz.ac.at!nmueller@hyper Tue Apr 26 05:53:56 1994 Date: Tue, 26 Apr 1994 03:56:41 -0400 To: hyperchem@hyper.com From: jk.uni-linz.ac.at!nmueller@hyper.uucp (Norbert Mueller) Subject: hyperchem suggestions HyperChem (referring to version 3.0 on PC, 2.0 on SGI) is a very appealing package. While using it in actual research work we have found a few rough edges which deserve - IMO - a little consideration. Some have already been mentioned here. Let me add another. Some results (the most annoying case for us are electronic spectra) are presented in modal dialog windows and there is no way, except for a screen dump, to get a hardcopy. I would suggest - obviously a I am late for version 4.0 - that any of these windows provides a way to save its contents not just as a screen dump but as text and/or a meta file. Still my biggest complaint: I cannot understand why the SGI platform has been abandoned. ---------------------------------------------------------------- Norbert Mueller - Institut f. Chemie, Johannes Kepler University A-4040 Linz Austria e-mail: NMUELLER@jk.uni-linz.ac.at (POP3 - Eudora - preferred) Norbert@soft.uni-linz.ac.at (MacPost - for personal messages) norbertm@convex.edvz.uni-linz.ac.at (unix - for BIG files) ________________________________________________________________ "True science teaches, above all, to doubt, and to be ignorant." Miguel do Unamuno ________________________________________________________________________ From uunet.ca!xantia.caspur.it!mencarel@hyper Tue Apr 26 10:50:45 1994 Date: Tue, 26 Apr 1994 10:31:07 -0400 From: Paolo Mencarelli To: hyperchem@hyper.com Subject: Electrostatic interactions calculated by using atomic charges to: hyperchem users group date: 26 April 1994 I am using HyperChem release 3, on a 486DX PC. When I carry out molecular mechanics calculations with the MM+ force field using the option of calculate electrostatic interactions by atomic partial charges (computed by a previous MO semiempirical calculation), I obtain strange results: 1) The total electrostatic energy is completely different from that obtained using the bond dipoles interactions. However this is quite expected and does not puzzle me! 2) Looking at the atom pair electrostatic interactions reported in the chem.log file, I noticed that: a) the assigment of atom types is wrong and it is not consistent with the atoms present in the molecule; b) the nergy value reported for each atom pair i, k does not correspond to that obtained with the formula indicated in the log file; c) the sign of the energy is not always consistent with the sign of the charges: i.e. sometimes the enrgy is positive for interaction of opposite sign charges or negative for charges with the same sign. This seems to happen even with very sinple molecules like 2-fluoroethanol. Does anyone have an answer tot the problem(s)? Thank you Dr. Paolo Mencarelli Dipartimento di Chimica Universita' La Sapienza Roma - ITALY E-Mail: mencarel@xantia.caspur.it ________________________________________________________________________ From slee@hyper Tue Apr 26 14:22:23 1994 From: slee@hyper.uucp (Thomas Slee) Subject: Re: Electrostatic interactions... To: mencarel@xantia.caspur.it (Paolo Mencarelli) Date: Tue, 26 Apr 1994 14:20:23 -0400 Cc: hyperchem@HYPER.COM Paolo Mencarelli writes... | When I carry out molecular mechanics calculations with the MM+ force field | using the option of calculate electrostatic interactions by atomic partial | charges (computed by a previous MO semiempirical calculation), I obtain | strange results: | | 2) Looking at the atom pair electrostatic interactions reported in the | chem.log file, I noticed that: | a) the assigment of atom types is wrong and it is not consistent with | the atoms present in the molecule; | b) the energy value reported for each atom pair i, k does not | correspond to that obtained with the formula indicated in the | log file; Dr. Mencarelli has indeed discovered a bug in HyperMM+. Fortunately, this is a printing bug only, so the final energies and gradients are being computed properly. The bug in printing the electrostatic energies is as follows: 1. Types are labelled incorrectly. Note that atom types are not employed in the calculation of the electrostatic energy. If you look in the MMPTYP.TXT file you will see that each type is offset by one row. 2. The "energy" reported is an accumulated total, not the interaction energy between the specified pair of atoms. The interaction energy is the difference between the value printed and the value reported for the previous pair. This covers all the problems mentioned. Tom Slee -- Tom Slee, Hypercube Inc., 419 Phillip St., Waterloo, Ont., Canada, N2L 3X2 E-Mail: slee@hyper.com, Tel. (519)725-4040, Fax. (519)725-5193.