From owner-hyperchem@hyper.com Thu Aug 3 18:23:28 1995 Date: Thu, 03 Aug 1995 19:15:58 +0000 (GMT) From: Paddy Kane <94970459@vax1.dcu.ie> Subject: distortions in optimizations To: hyperchem@hyper.com Cc: support@hyper.com Hi, I've asked this question before but since I did no explain my problem very well and since I still haven't managed to find a satisfactory solution please permit to ask the question again. Upon examining some optimizations I sometimes find that compounds containing benzene rings or fused benzene rings have undergone SEVERE distrtions. For example: 1) one or more of the bonds to the ring are bent out of the plane of ring, sometimes by as much 40 degrees. With some compounds I can highlight the ring and MODEL build (from the EDIT menu) that ring. In other compounds the distorted ring(s) are bonded to other rings to form a macrocycle and selecting one of the distorted rings and trying to model build it results in an error message sating that the selection is inappropriate. In this case I have resorted to continuously setting and ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 3 20:11:23 1995 Date: Thu, 03 Aug 1995 20:47:25 +0000 (GMT) From: Paddy Kane <94970459@vax1.dcu.ie> Subject: Distortions in Optimisations To: hyperchem@hyper.com Cc: support@hyper.com Hi, This is a repeat of a message which was sent incomplete to the list earlier. I sincerely apologise for any inconvenience caused. I have asked this question before but since I never found a satisfactory answer please permit me to ask the it again. I sometimes find that optimisations do not converge as a result of SEVERE distortions. If I can remove the distortions by adjusting the geometry, the optimisation converges and does so reasonably quickly. The sort of distortions I am referring to are as follows: 1) compounds containing benzene rings are distorted so that one or more of the bonds from such a ring are bent out of the plane of the ring, sometimes by as much as 40 degrees. I can eventually bring the offending bond back into the plane of the offended ring by continuously resetting the bond angles which the bond makes with the ring to 120 degrees. 2) fused benzene rings are bent out of the plane the plane they were in originally at the start of the optimisation, sometimes by as much as 90 degrees. By MODEL building the rings I can make them planar again if the distorted rings are at the end of a side chain. 3) benzene rings which are bonded to other rings by CH2 groups to form macrocycles are sometimes distorted out of their plane by as much as 90 degrees. In this case the MODEL building method will work only if I break bonds and reform them again after the model building. Am I missing something or do I have to continue to use these cumbersome methods? Regards, Paddy. Paddy Kane School of Chemical Sciences Dublin City University Glasnevin Dublin 9 Ireland e.mail: 94970459@vax1.dcu.ie ________________________________________________________________________ From owner-hyperchem@hyper.com Fri Aug 4 11:41:27 1995 From: "m.taylor" To: hyperchem@hyper.com Date: Fri, 4 Aug 1995 09:33:18 +0000 Subject: those hardware locks again Reply-To: m.taylor@pra.org.uk >> Does anyone else have problems with their hardware lock? Any solutions? I have had continual problems with running the hardware lock under Windows NT 3.1 and 3.5 Workstation, although it has always been OK under Windows. I had a lot of help from Hypercube in getting the configuration of the Rainbow driver right, but it still only worked with a delay of 15 seconds on every dongle-checking operation, with a bonus delay of 2 minutes on the first such operation! I have had a complete new motherboard on the machine I use (32MB Pentium 90), plus upgrading from NT 3.1 to 3.5, without any improvement. Recently I tried the "new" Rainbow driver passed on by Hypercube, but this crashed my system so hard I had to do a complete reinstallation of NT Workstation. I have been told on the authority of Rainbow that the hardware lock "either works immediately or not at all" under NT - so that in my case, the fact that it works after 15sec - 2 mins implies...just what does it imply? What else can I do? In desperation, Hypercube, is there any possibility my dealer in the UK would be allowed to swap my existing dongle for a fresh one, just in the remote possibility that the wretched thing may have some physical flaw? My sympathies to everyone else who suffers from hardware locks! Mike Taylor =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=* Mike Taylor Paint Research Association 8 Waldegrave Road TEDDINGTON TW11 8LD ENGLAND Tel: +44 (181) 977 4427 Fax: +44 (181) 943 4705 =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=* ________________________________________________________________________ From polowin Tue Aug 8 12:20:16 1995 Date: Fri, 4 Aug 95 14:31:08 -0400 From: polowin (Joel Polowin) To: Jeff Rasimas Subject: RE: those hardware locks again Cc: hyperchem@www > From: Jeff Rasimas > Subject: RE: those hardware locks again > Date: Fri, 04 Aug 95 13:20:00 E > > I have also had problems with my hardware lock under NT 3.51. It worked > well under NT 3.5, but now I get this error message "Communications Error" > which locks up Hyperchem, and often the entire system. > > Any ideas or solutions (I thought that I was the only one, my mailer's also > been sporadic so I've not gotten this thread till now).!? I just spoke with a tech-support fellow at Rainbow. His suggestion is that if NT 3.51 was installed over NT 3.5, that might have "deconfigured" the port driver. His recommended deleting the old copy of the port driver and re-installing it. Rainbow has a new version of the Rainport driver which they're going to be sending me. He told me that their new installation procedure is able to configure the port driver correctly without requiring the use to extract hardware-address information manually. 98 or 99% of the time. I suppose that that's an improvement. As soon as I receive the thing I'll make it available from our WWW and ftp sites, and I can E-mail the package to those who don't have WWW or ftp access. Stay tuned... Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Tue Aug 8 12:24:19 1995 Date: Fri, 4 Aug 95 15:07:31 -0400 From: polowin (Joel Polowin) To: hyperchem@www Subject: HyperChem Workshops Announcement >From Dragan Vuckovic at Hypercube: Dear CCLers, Hypercube, Inc. is pleased to announce that it is organizing a series of workshops on molecular modeling, computational chemistry, and the use of HyperChem, ChemPlus and HyperNMR. It is offering a three-day introductory (HyperChem's GUI, Molecular Mechanics, Approximate and Semi-empirical MO methods, ab initio MO methods, Molecular Dynamics, Molecular Spectroscopy and an introduction to QSAR. The course will assume some basic knowledge of chemistry and a very moderate knowledge of PC computers and Microsoft Windows) and advanced (Molecular Modeling techniques, HyperChem's GUI - Advanced Features, QSAR and QSPR, Systems with Periodic Boundaries, Molecular Dynamics and Monte Carlo Methods, Thermodynamics and Kinetics, Script Commands and Customization, and linking HyperChem to other programs - Visual Basic and Excel) courses. Workshops on site are also available, as well as one-day courses on specific issues. Please contact Hypercube, Inc. (vuckovic@hyper.com) or HyperChem dealers for more details on workshops, or visit Hypercube's home page (URL: http://www.hyper.com). ________________________________________________________________________ From owner-hyperchem@hyper.com Tue Aug 8 12:31:23 1995 Date: Fri, 4 Aug 95 14:31:08 -0400 From: polowin (Joel Polowin) To: Jeff Rasimas Subject: RE: those hardware locks again Cc: hyperchem@www > From: Jeff Rasimas > Subject: RE: those hardware locks again > Date: Fri, 04 Aug 95 13:20:00 E > > I have also had problems with my hardware lock under NT 3.51. It worked > well under NT 3.5, but now I get this error message "Communications Error" > which locks up Hyperchem, and often the entire system. > > Any ideas or solutions (I thought that I was the only one, my mailer's also > been sporadic so I've not gotten this thread till now).!? I just spoke with a tech-support fellow at Rainbow. His suggestion is that if NT 3.51 was installed over NT 3.5, that might have "deconfigured" the port driver. His recommended deleting the old copy of the port driver and re-installing it. Rainbow has a new version of the Rainport driver which they're going to be sending me. He told me that their new installation procedure is able to configure the port driver correctly without requiring the use to extract hardware-address information manually. 98 or 99% of the time. I suppose that that's an improvement. As soon as I receive the thing I'll make it available from our WWW and ftp sites, and I can E-mail the package to those who don't have WWW or ftp access. Stay tuned... Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Tue Aug 8 12:46:42 1995 Date: Fri, 4 Aug 95 14:21:47 -0400 From: polowin (Joel Polowin) To: hyperchem@www Subject: Re: Distortions in Optimisations > Date: Thu, 03 Aug 1995 20:47:25 +0000 (GMT) > From: Paddy Kane <94970459@vax1.dcu.ie> > Subject: Distortions in Optimisations > > I sometimes find that optimisations do not converge as a > result of SEVERE distortions. If I can remove the > distortions by adjusting the geometry, the optimisation > converges and does so reasonably quickly. If you start with a seriously distorted structure, then often what comes out of a structural optimization is a local minimum, not a global one. Particularly when the desired structure is strained, or is something like a macrocycle with lots of projections and different ways of twisting around, you will need to tweak things manually. The computer sees things at a very local level -- it moves each atom in a direction that leads to a lower energy. It can't see the big picture: that if a large set of atoms is twisted around in a particular way then two sections will no longer be colliding, for example. > Am I missing something or do I have to continue to use > these cumbersome methods? I would suggest that you create a "library" of basic structures -- the ring backbones and common variants saved as files. Then you can read those back in and elaborate upon them, adding side chains and other modifications as necessary. If you're starting with structures that have at least reasonable conformations, then optimizing them after modification should run smoothly. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Tue Aug 8 18:20:04 1995 Date: Tue, 8 Aug 95 13:30:00 -0400 From: polowin (Joel Polowin) To: Jeff Rasimas Subject: Re: dongle on NT3.51 Cc: hyperchem@www > From: Jeff Rasimas > Subject: dongle on NT3.51 > Date: Fri, 04 Aug 95 16:02:00 E > > I've been checking the rainbow dongle on a new P5-120 running NT3.51. I > have the dongle installed and Hyperchem 4.0 runs fine for a bit. For > example, if I try to minimize 2 water molecules using MM+, the computation > will optimize the geometry for 1934 cycles (tried it 4 times, crashes at the > same point). This computer was purchased to run Hyperchem, and it'll be a > waste if i can only compute for 90 seconds at a time! > Hyperchem 4.0 ran fine on NT3.5 on a 486-66. > > Any suggestions? I really HATE this dongle, it's a PITA! First suggestion: Try the new rainport driver. I've just installed it on our home page (http://www.hyper.com); you can get to it from "What's New" or from the second FAQ page (http://www.hyper.com/Support/Faq/faq2.html); you can download it directly as ftp://ftp.hyper.com/support/rainport.exe . By ftp, you can get it from ftp.hyper.com in the "support" directory as rainport.exe . I can, of course, send copies of the new program by E-mail on request. To extract the contents of the file, transfer it to a directory and enter rainport -d -- the "-d" option recreates the directory structure in the file, and is necessary to install the driver correctly. The new version, unlike the old, is supposed to be able to configure itself without the user having to determine hardware addresses manually. If this doesn't help, please let me know and I'll whine at Rainbow some more. Second suggestion: see if running NT 3.5 vs. 3.51 makes a difference; either install 3.5 on your new machine to see if HyperChem starts working there, or 3.51 on your old machine to see if it stops. Third suggestion, and please don't think that I'm being snide or sarcastic: Our software runs quite nicely on Pentiums under Windows. With Windows we don't have to worry about using extra software like the port driver to get the software to communicate with the lock. I know that NT offers a number of nice features, but unless you're doing a lot of multi-tasking, it shouldn't offer significantly better performance for HyperChem. Regards, Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Tue Aug 8 21:55:32 1995 Date: Wed, 9 Aug 1995 08:39:05 +1000 X-Sender: bromrd92@octarine.adfa.oz.au To: hyperchem@hyper.com From: bromrd92@octarine.cc.adfa.oz.au (Marcus Bromet) Subject: Returned mail: warning: cannot send message for 4 hours Hello Joel and HyperChemists, I have performed a semi-empirical calculation on a molecule (which takes approximately 24 hours to complete). At this point I can view contour and orbital plots as per the COMPUTE menu (HyperChem V4.0 for Windows). However if I save the molecule, turn the computer off, and come back the next day to continue work, I cannot reload all the data: i.e. I can reload the structure, view the atom types labelled with charges calculated by the semi-empirical calculation, but I can no longer do a contour or orbital plot. It seems pointless to have to redo a single point semi-empirical, every time I wish to do a contour plot - so I think I am missing something regarding the saving of semi-empirical data that I cannot find in the manuals. Please assist if possible. With Thanks Marcus Bromet, Chemistry Department, University College (UNSW) Australian Defence Force Academy, CANBERRA ACT 2600 Australia voice: +61 6 2688943(w)/2688922(h) fax:+61 6 2688002 e-mail:bromrd92@octarine.adfa.oz.au WWW:http://apamac.ch.adfa.oz.au/10NC/program/divisions/inorg/molrec/bromet/ ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 10 02:41:32 1995 Date: Wed, 9 Aug 95 17:51:50 -0400 From: polowin (Joel Polowin) To: hyperchem@www Subject: Re: Ab initio performance > From: "MORTEN LANGGaaRD - BIOKEM" > Date: Thu, 27 Jul 1995 11:32:24 +0100 It's taken us a bit longer than it really should have to do some testing in response to Dr. Langgaard's comments comparing the timing and comparisons of HyperChem and Gaussian 92. HyperChem is actually faster than Gaussian 92 for SCF calculations and Langgaard's results represent limitations in HyperChem geometry optimizations that we hope to address at some future point. The computer used for doing these timing and comparisons was a Pentium 90 with 16 MB RAM and 1 GB hard disk. The following tests were performed by Yufei Guo: 1) Benzene: Geometry from the HyperChem model builder. Single Point Calculation Basis set HyperChem G92 STO-3G 59 sec 1 min 35 sec 3-21G 5 min 24 sec 5 min 50 sec 6-31G 6 min 10 sec 6 min 37 sec The way to record these times are: We start timing HyperChem calculation after issuing Compute/Single Point and end timing when the HyperChem calculation has done. The Gaussian 92 times are from their .out file. 2) Ethanol: Geometry is first optimized by AM1. Geometry Optimization. Basis set HyperChem G92 STO-3G: Calc. Time 9 min 28 sec 6 min 20 sec # of SCF calc. 36 5 Final Grad 0.066 kcal/mol/A 0.095 kcal/mol/A 3-21G: Calc. Time 37 min 53 sec 12 min 14 sec # of SCF calc. 38 5 Final Grad 0.044 kcal/mol/A 0.055 kcal/mol/A Comments: HyperChem is faster for single point calculations than Gaussian 92. However, Gaussian 92 is faster than HyperChem for geometry optimizations, principally because HyperChem doesn't calculate second derivatives analytically. Also, the optimization algorithms in Gaussian 92 (which use second derivatives) are better than those in HyperChem (which use first derivatives). HyperChem's speed could be improved (for semi-empirical calculations as well) by introducing better optimization algorithms and analytical second derivatives. These are goals for future releases of HyperChem. Remember that this is our first ab initio offering. While HyperChem doesn't yet compute second derivatives analytically, its first derivatives are probably faster than those of Gaussian 92 (see the above benchmarks) so that HyperChem calculations such as those involving chemical reaction trajectories and molecular dynamics would be faster than Gaussian 92 (if these kinds of calculations were available in Gaussian). HyperChem's Direct SCF is slower than its non-direct SCF method, which stores and does not recompute the two-electron integrals. This is a simple space/time trade-off, and the results are unsurprising. > Here is some notes on "ab initio" performance in HyperChem. > > I received my HyperChem 4.5 upgrade for a few weeks ago and was very > curious to see how well the new ab initio part is implemented. As > usual with HyperChem I find that the user interface is almost > perfect for the non specialist user - everything you really want is > there - just a click away. Applause, applause! > > To check the new Gauss engine, I decide to run some simple tests on > a little well known molecule - Ethanol. > For comparison I ran the same test with the original Gaussian 92 for > Windows program from Gaussian, Inc. The results listed below, are all > started from the same input structure (AM1 optimized gnorm=0.1). > Termination condition was set to gnorm=0.1 in HyperChem and the > default "tight" in G92W. It is not straight forward to compare these > two criteria, but I guess that they are very close, since HyperChem's > gnorm=0.1 will be satisfied after just a few cycles on a G92W > optimized structure (this will always be necessary due to rounded > figures). The computer used for these calculation is a AMD DX4-100MHz > with 32Mb ram, 256kb cache, and a reasonable fast Western Digital > 730Mb HD. > > Geometry optimization of ethanol: > > HyperChem, STO-3G, direct SCF, 32 min. > HyperChem, STO-3G, non-direct SCF, 16 min. > G92W, STO-3G, direct SCF, 6 min. 35 sec. > > HyperChem, 3-21G, direct SCF, 1 H. 30 min. > HyperChem, 3-21G, non-direct SCF, 45 min. > G92W, 3-21G, direct SCF, 8 min. 33 sec. > G92W, 3-21G, non-direct SCF, 10 min. 21 sec. > > HyperChem, 6-31G*, non-direct SCF, 4 H. 33 min. > G92W, 6-31G*, direct SCF, 1 H. 12 min. > > Comments: > As it follows from these figures, it is exactly twice as > expensive to run direct SCF with HyperChem - this is strange! > With G92W it doesn't really matter and my experience is, that the > direct method often is the fastest. There is not mush logic in this > and as the G92W manual says "you have to try". By using the direct > method you don't have to worry about disk space. > > Why is HyperChem about 4-6 times slower than G92W? > > Well, I don't know - maybe I do something wrong! But I have tried > the following changes from the default settings: > Raffenetti integral format - No effect at all! (G92W uses these as default) > Buffer size increased to 10000 - no effect! (G92W uses 8192 words for these > calculations) > > A general change I have made is to use INDO MO initial guess instead > of the default Core Hamiltonian (INDO is also used by G92W). My short > experience is that the Core Hamiltonian often fails totally and > causes corrupted structures (try it on Ethanol and you may see the > end of this molecule - whooh). Any explanation? > INDO guess's are fast and seem to work every time. > > By default G92W has "Two-electron integral symmetry turned on" > (keyword SYMM) which might speed up the calculation. The symmetry can > be turned off by the keyword NOSYMM, but this only causes slightly > longer computational times (about 5%). I don't know what HyperChem > use regarding these "two electron" symmetries. > > Is it so bad that the Gauss.exe code simply is slow or is there some > good explanation for these figures? > Do we have to wait for release 5 (and Win9?)? ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 10 19:00:48 1995 Date: Thu, 10 Aug 95 15:35:47 -0400 From: polowin (Joel Polowin) To: bromrd92@octarine.cc.adfa.oz.au (Marcus Bromet) Subject: Re: Retrieval of semi-empirical results Cc: hyperchem@www > Date: Wed, 9 Aug 1995 08:39:05 +1000 > From: bromrd92@octarine.cc.adfa.oz.au (Marcus Bromet) > > Hello Joel and HyperChemists, > I have performed a semi-empirical calculation on a molecule > (which takes approximately 24 hours to complete). At this point I can view > contour and orbital plots as per the COMPUTE menu (HyperChem V4.0 for Windows). > However if I save the molecule, turn the computer off, and come back > the next day to continue work, I cannot reload all the data: > i.e. I can reload the structure, view the atom types labelled with > charges calculated by the semi-empirical calculation, but I can no longer do > a contour or orbital plot. > > It seems pointless to have to redo a single point semi-empirical, > every time I wish to do a contour plot - so I think I am missing something > regarding the saving of semi-empirical data that I cannot find in the manuals. I'm afraid not. It's something that was talked about before the new release, but we weren't able to get it implemented in time. The release for SGI will be able to store this information in a limited way, I'm told, though still not in a complete form. The Molecule Presentations module of ChemPlus is able to store the 3D grid data in a file, and bring it back. This sort of information should be saveable and retrievable in a HIN file in the next version of HyperChem. (Not soon! We just released a version!) It's high on the wish list. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 10 19:28:27 1995 Date: Thu, 10 Aug 95 15:00:28 -0400 From: polowin (Joel Polowin) To: bromrd92@octarine.cc.adfa.oz.au (Marcus Bromet) Subject: Re: Pseudo Atoms (?) Cc: hyperchem@www Hi, Marcus. > Its been some time since I have asked you any questions so I thought > I would get you with a double whammy! (thats what we call them in Australia > - it means two at once). The phrase is used here, too. I don't know its origin... > I have two molecules, both of whom have centres of mass which are > hollow (or are cavities). Firstly I want to define a centre of each > molecule and see it as a visible point (i.e a psuedo atom I think is what > they are called). I would like HyperChem do that for me so that the pseudo > atom in each hollow cavity does not effect either molecule or its interactions. > > QUESTION 1: Can HyperChem define the centre of mass of a molecule, and do so > by defining it as a pseudo atom which does not effect the properties of the > molecule? Sort of, but not usefully for this problem, I think. You can define a single named selection as POINT; then when you use a couple of the dialogue boxes for things like Tethering, you can get the Cartesian coordinates of POINT -- the location of the centre of mass. But if you try to tether an atom to POINT, it gets tied to the Cartesian coordinates of POINT at the time of tethering, not to the floating location of POINT itself. Now, you could have several named selections, and you could run a script or something so that a structure was selected by name and then the name POINT applied to it -- replacing the previous POINT definition. You could then get the coordinate information about it. But this wouldn't give you information about the distance between the two, not directly at least. You could run a macro and calculate the distance from the several pieces of data. What you *can* do is to set up a couple of "dummy atoms" or "pseudo-atoms" in a different way. In molecular mechanics calculations, an atom which is not connected to any structure and whose van der Waals epsilon parameter has been set to 0, and which is uncharged, does not interact in any way with other structures. In semi-empirical or ab initio calculations, the "lone pair" atoms behave the same way. If you add one of these atoms to a system, you can add Restraints in such a way as to keep the dummy more or less where you want it with respect to the rest of the system. For example, if you have a 5-membered symmetrical planar ring, you could put a dummy atom 'x' near its centre, and add restraints to keep angles 1-x-3 and 1-x-4 at 144', and 3-x-4 at 72'. This keeps the dummy at the centre of the ring (unless the ring itself becomes seriously distorted), and -- importantly -- does not exert any forces on the ring atoms when the dummy has moved to the proper location. By turning the restraints off and running a single-point calculation, you'll get the same results in all respects as though the dummy weren't there at all. This has obvious applications for constraining the conformation of a structure -- say, the orientation of a pi-coordinated system with respect to a metal atom. With an ethylene complex, one would add a dummy near the centre of the C=C, and restrain the C-x-C angle to be 179.9999' (can't make it exactly 180') and one of the C-x distances to be about half of the expected C-C length. Again, when the dummy is correctly placed this gives no net forces on the C atoms. Then one could restrain the C-x-M angle for M a coordinated metal atom, or the C-x-M-L torsion for another ligand. In your case, you could set up one dummy atom for each molecule. You would have the problem of defining things so that the dummies represented the positions that you wanted. I'd suggest that for the dioxan you might try an O-x-O angle of 180' and some appropriate value for one O-x; some set of three angles, as with the 5-membered ring above, might be found for the cyclodextrin. > Then I want to use HyperChem to measure the distance between the two > molecules by measuring the distance between the two pseudo atoms (simply by > having the two pseudo atoms selected). I then want to do a minimisation > such that as one molecule approaches the other the distance of closest > approach is recorded in the log file. Further if the two molecules interact > so that one goes inside the other (i.e one forms an inclusion complex with > the other) the pseudo atoms so not interfere in any way, but still record > the distance of the approach. This sounds like a dynamics problem. If you run Molecular Dynamics on this system with all initial velocities set to 0 (and check the Restart option), with a sufficiently long running time you should be able to get a system minimization out of this. (Use a low bath temperature, such as 0.) This lets you monitor something like a bond distance: apply a name to the two dummy atoms, and then use the Dynamics Averages functions to track the distance through the course of the simulation. Regards, Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Wed Aug 16 18:48:06 1995 Date: Wed, 16 Aug 1995 19:40:56 +0000 (GMT) From: Paddy Kane <94970459@vax1.dcu.ie> Subject: RUTHENIUM To: HYPERCHEM@HYPER.COM, CHEMISTRY@infomeister.osc.edu Hi HyperChemers CCL'ers A colleague of mine needs parameters for Ruthenium so that she can molecular mechanics calculations in HyperChem with the MM+ force field. If you can supply such information for any oxidation state, we would be much obliged. Paddy. Paddy Kane Dublin City University Ireland 94970459@vax1.dcu.ie ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 17 11:07:05 1995 From: simonk@matthey.com Date: Thu, 17 Aug 1995 06:26:43 -0400 To: hyperchem@hyper.com Subject: Parameters for Platinum Group Metals Hi All, With regards to Paddy Kane's recent e-mail for MM+ parameters for Ru, are there decent parameters, semi-empirical or molecular mechanics which go further than those presently in Release 4.5 for the other Platinum Group Metals? Thanks in advance, Keith ********************************************************************* =20 Dr. Keith Simons TEL: +44-(0)1763-253126 Senior Development Chemist FAX: +44-(0)1763-253516 Johnson Matthey E-Mail simonk@matthey.com Process & ElectroCatalyst Development Orchard Road Royston Herts. SG8 5HE UK =20 ********************************************************************** ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 17 15:41:28 1995 Date: Thu, 17 Aug 95 11:53:25 -0400 From: polowin (Joel Polowin) To: simonk@matthey.com Subject: Re: Parameters for Platinum Group Metals Cc: hyperchem@www > From: simonk@matthey.com > Date: Thu, 17 Aug 1995 06:26:43 -0400 > > With regards to Paddy Kane's recent e-mail for MM+ parameters for Ru, are > there decent parameters, semi-empirical or molecular mechanics which go > further than those presently in Release 4.5 for the other Platinum Group > Metals? I'm not sure what you mean by "go further". None of the semi-empirical methods in the current versions of HyperChem can handle elements beyond xenon -- that's high on the wish-list for future development. We don't have references for published parameters that give better results for the current set of atoms, but we'd certainly like to hear about them if there are any. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 17 18:03:19 1995 Date: Thu, 17 Aug 95 14:59:41 BS3 From: CCATF%BRLNCC.BITNET@VTBIT.CC.VT.EDU To: hyperchem@HYPER.COM Hyperchem: testing, please acknowledge. dr. c. a. taft ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 17 18:47:02 1995 Date: Thu, 17 Aug 95 16:06:01 BS3 From: CCATF%BRLNCC.BITNET@VTBIT.CC.VT.EDU To: hyperchem@HYPER.COM hyperchem: l testing, please acknowledge receipt. ________________________________________________________________________ From owner-hyperchem@hyper.com Fri Aug 18 19:03:45 1995 Date: Fri, 18 Aug 95 16:20:28 BS3 From: CCATF%BRLNCC.BITNET@VTBIT.CC.VT.EDU To: hyperchem@HYPER.COM hyperchem: how do you fix atoms in a plane for visualization of other atoms and molecules not in this plane. ANTHONY ________________________________________________________________________ From owner-hyperchem@hyper.com Mon Aug 21 16:17:22 1995 Date: Mon, 21 Aug 95 10:34:27 -0400 From: polowin (Joel Polowin) To: CCATF%BRLNCC.BITNET@VTBIT.CC.VT.EDU Subject: Molecule alignment Cc: hyperchem@www > Date: Fri, 18 Aug 95 16:20:28 BS3 > From: CCATF%BRLNCC.BITNET@VTBIT.CC.VT.EDU > > hyperchem: how do you fix atoms in a plane for visualization > of other atoms and molecules not in this plane. I'm not quite sure what you're asking. If you wish to rotate a structure so that a set of atoms are in a plane perpendicular to the line of view, you can select the three (or more) atoms of the plane, and use Edit/Align Molecules to align the tertiary axis with Z axis. (Alternately, you could align the primary and secondary axes with X and Y or vice versa.) This aligns the structure with the viewer coordinate axes. If, before you do this, you use Edit/Align Viewer to align the viewer with the Z axis of the molecular coordinate system, you can align the molecular plane with the coordinate plane of the molecular coordinate system. Another useful trick is to save the structure in Z-matrix format and then read it back in. This places the first atom in the structure at coordinates (0,0,0), the second along the positive X axis, and the third in the X plane. If this doesn't answer your question, would you please clarify what you mean by "fixing atoms in a plane"? Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. ________________________________________________________________________ From owner-hyperchem@hyper.com Mon Aug 21 16:56:08 1995 From: "Showalter, Steven K." To: HyperChem ListServer Subject: Windows 95 Date: Mon, 21 Aug 95 10:34:00 MDT Joel, How well will HyperChem work with Windows 95? Have you guys tested it? Are there any kludges we should be aware of? Any problems with the Rainbow lock? Steve Showalter skshowa@sandia.gov skshow@rt66.com http://www.rt66.com/skshow ________________________________________________________________________ From owner-hyperchem@hyper.com Tue Aug 22 17:09:38 1995 Date: Tue, 22 Aug 95 18:15 +0200 From: To: amber@cgl.ucsf.edu, anneal@chaco.com, c2-l@msi.com, cache@pacificu.edu, charmm-bbs@emperor.harvard.edu, chem-com@mailbase.ac.uk, watoc@ic.ac.uk, chemconf@umdd.umd.edu, chemistry@infomeister.osc.edu, chem-mod@mailbase.ac.uk, hyperchem@hyper.com, isisforum-l@mdi.com, mmodinfo@uoft02.utoledo.edu, quanta-l@msi.com, pharm@dmu.ac.uk Subject: WATOC96 WATOC '96 - CONFERENCE ANNOUNCEMENT WORLD ASSOCIATION OF THEORETICALLY ORIENTED CHEMISTS 4TH CONGRESS, JERUSALEM, ISRAEL, 7 - 12 JULY 1996 The purpose of the Fourth WATOC World Congress is to bring together scientists interested in the development of various theoretical methods, including molecular modeling, and their application to organic, inorganic, organometallic, biochemical, and biomedical problems. The WATOC Congress also is intended to attract experimentalists who may wish to employ computational methods in their research. Thirteen Plenary lectures, 40 invited talks and about 50 oral presentations, as well as two poster sessions, are scheduled. All presentations will be published in a special issue of the Journal of Molecular Structure (Theochem), as is the regular practice for all the plenary and invited lectures of WATOC conferences. About a third of the Congress will be dedicated to the development of theoretical methods, a third to applications in organic, inorganic, and organometallic chemistry, and a third will deal with biochemical/biomedicinal theory and applications. Plenary lectures will be delivered by the following distinguished scientists: T. Blundell, D. Cremer, W.F. van Gunsteren, K.N. Houk, P. Kollman, W. Kutzelnigg, K. Morokuma, J.A. Pople, H. Schwarz, S. Shaik, W. Thiel, F. Weinhold, and H. Weinstein. Invited speakers who already have agreed to participate include: D. Avnir, M.V. Basilevsky, W.T. Borden, T. Clark, D.S. Dudis, R. Elber, G. Frenking, J. Gauss, J. Gerrat, P.M.W. Gill, P.C. Hiberty, S. Hoz, A. Itai, U. Kaldor, M. Karni, W. Koch, D. Kost, P. Lam, J.P. Malrieu, K. Merz, K. Muller, S. Nagase, E. Osawa, A. Pross, P. Pyykko, L. Radom, G. Richards, H. Rzepa, A. Sali, J. Sauer, H.F. Schaefer, H.B. Schlegel, P.v.R. Schleyer, B. Schoichet, P. Schwerdtfeger, G. Scuseria, J. Skolnick, R. Sustmann, A. Warshel In addition to the scientific sessions, attractive social events are planned, as well as a program for accompanying persons. Pre- and post-conference tours of Israel (with possible inclusion of Jordan and Egypt) will be organized. Visas are guaranteed to all scientists who register. The Congress will take place at the Holiday Inn Crowne Plaza Hotel in Jerusalem which will be the most convenient location for the participants to stay. Several other hotels, located in the vicinity (within walking distance), are available at prices ranging from about $75 to $150 for a double room (including an Israeli, i.e., large breakfast, service charges and taxes -1995 prices). Prices for single occupancy are $70-$120 in the same hotels. A small number of rooms will be reserved in the student dormitories ($25 per room/night - with no breakfast included). Jerusalem may be reached easily from Ben-Gurion Airport near Tel-Aviv (ca. 45 min.). This is the regular airport for all international flights to Israel. Taxis offer excellent transportation from the airport to Jerusalem (about $45 for single, $10 for shared occupancy). Bus service is available as well for ca. $5. In order to receive circulars, we ask prospective participants to respond to this announcement by sending us their full names, address, tel. and fax numbers and e-mail address. Also, please indicate if you have access to World-Wide-Web (for future registration, hotel reservation, abstract transfer, etc purposes). For responses and/or further information, please contact: Prof. Amiram Goldblum, Chairman of WATOC '96 Tel. 972-2-758701, FAX: 972-2-410740 e-mail: watoc96@vms.huji.ac.il OR Prof. Yitzhak Apeloig, Program Chairman of WATOC '96 Tel. 972-4-293727, FAX: 972-4-233735 e-mail: chdean@techunix.technion.ac.il WATOC WWW server: http://www.ch.ic.ac.uk/watoc/ _______________________________________________________________ Amiram Goldblum Prof. of Pharmaceutical Chemistry Phone 972-2-758701 Hebrew University of Jerusalem FAX 972-2-410740 School of Pharmacy email amiram@vms.huji.ac.il Jerusalem 91120 ISRAEL ________________________________________________________________ ________________________________________________________________________ From owner-hyperchem@hyper.com Wed Aug 23 09:38:17 1995 Date: Wed, 23 Aug 1995 17:57:28 +1000 From: tony_hughes.pharmacology@muwaye.unimelb.edu.au (Tony Hughes) Subject: ...no subject... To: hyperchem@hyper.com Dear Hyperchem Users, I've got the monomer coordinates for a dimeric protein, and wish to obtain a structure for the dimer. I've recently received a matrix from the crystallographer who did the structure, that should give me the other monomer. The matrix looks like this: ASYM 0.50000 -0.86603 0.0000 -0.86603 -0.50000 0.0000 0.00000 0.00000 -1.0000 28.2401 48.9132 30.3739 Is ther some way I can use this matrix to transform my protein monomer in Hyperchem to give me the dimer I want? I've tried simply editing this matrix into the viewing transform field of the .HIN, without success. Thanks if you can help. Cheers, Tony Hughes ________________________________________________________________________ From owner-hyperchem@hyper.com Wed Aug 23 15:43:48 1995 Date: Wed, 23 Aug 95 12:06:12 -0400 From: polowin (Joel Polowin) To: tony_hughes.pharmacology@muwaye.unimelb.edu.au (Tony Hughes) Subject: Re: HyperChem transformation Cc: hyperchem@www > Date: Wed, 23 Aug 1995 17:57:28 +1000 > From: tony_hughes.pharmacology@muwaye.unimelb.edu.au (Tony Hughes) > Subject: ...no subject... > > I've got the monomer coordinates for a dimeric protein, and wish to obtain a > structure for the dimer. I've recently received a matrix from the > crystallographer who did the structure, that should give me the other > monomer. The matrix looks like this: > > ASYM > > 0.50000 -0.86603 0.0000 > -0.86603 -0.50000 0.0000 > 0.00000 0.00000 -1.0000 > 28.2401 48.9132 30.3739 > > Is ther some way I can use this matrix to transform my protein monomer in > Hyperchem to give me the dimer I want? I've tried simply editing this matrix > into the viewing transform field of the .HIN, without success. I'm not a crystallographer, but this matrix appears to me to involve a 3x3 rotation conversion (with a 60' angle) and a 1x3 translation. If I take a molecule, select one atom, and (with whole-molecule translation on) do Edit/Translate/Translate Selection to Origin, the molecule is moved so that that atom is at 0,0,0 . Save the structure as a .HIN file at this point. Unselect all. Edit/Rotate applied to Molecules, with a 60' rotation about Z followed by a 180' rotation about X, seems to give the desired transformation, with coordinates of the non-origin atoms being moved appropriately. Now you can select the origin atom again and use the Edit/ Translate/Translate Selection to 28.2401, 48.9132, 30.3739 ; this will translate the whole molecule. You can save this structure in a new file. Use a script command or macro to execute the script command translate-merged-systems no so that when you read in one file, and File/Merge in the other, the coordinates in the files are maintained. *IF* I am interpreting the matrix correctly -- and, again, I am not a crystallographer! -- this should give you the results you're looking for. It would probably be a good idea to check with a crystallographer for the proper interpretation of the matrix. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. a ________________________________________________________________________ From owner-hyperchem@hyper.com Sat Aug 26 19:27:16 1995 Date: Sat, 26 Aug 95 18:09:08 BS3 From: CCATF%BRLNCC.BITNET@VTBIT.CC.VT.EDU To: hyperchem@HYPER.COM hyperchem: SUPPOSE I HAVE A LARGE MOLECULE DENOTING A SURFACE. AND I WANT TO ATTACH A MOLECULE WATER FOR EXAMPLE TO THIS SURFACE WITHOUT MOVING THE POSITIN S OF LARGE MOLECULE, POSSIBLE ADJUST THE DISTANCE OF WATER TO LARGE MOLECULE WITHOUT AFFECTING THE COORDINATES POSITIoN OF LARGE MOLECULE? ANTHONY ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 31 06:40:49 1995 From: Massimo Trotta Subject: non-aqueos solvent box To: hyperchem@hyper.com Date: Thu, 31 Aug 1995 09:22:39 +0200 (MET DST) hallo everybody, any suggestion for building a solvent box with solvent different from water? has anybody done it? also what should I do exactly for retrieving information from the www page of hyperchem?: I think I read that there is a macro for rachamandran plot available there. Thanks for you help massimo ________________________________________________________________________ From owner-hyperchem@hyper.com Thu Aug 31 18:26:36 1995 Date: Thu, 31 Aug 95 15:34:40 -0400 From: polowin (Joel Polowin) To: Massimo Trotta Subject: Re: non-aqueos solvent box Cc: hyperchem@www > From: Massimo Trotta > Subject: non-aqueos solvent box > Date: Thu, 31 Aug 1995 09:22:39 +0200 (MET DST) > > hallo everybody, > any suggestion for building a solvent box with solvent different from water? > has anybody done it? I've done it a couple of times. Here's what we've got on the subject in the revised (for the SGI release) reference manual: Solvents other than water Pre-defined files for periodic boxes of solvents other than water are not distributed with HyperChem, but creating your own solvent boxes is relatively straightforward. To find the number of solvent molecules appropriate for your box, you will need the box volume (length x width x height) and the molecular weight and density of your solvent. Then: # molecules = (volume x density x Avogadro's constant) / molecular weight For example, for a periodic box 18.62 A in each dimension, filled with water at 25'C, the density is 1.00 g/cm3 and the molecular weight is 18.02 g/mol. Substituting into the above, and including the conversion from A to cm: # molecules = (18.62 A)^3 x (1 cm/10^8 A)^3 x 1.00 g/cm^3 x 6.023E23 / 18.02 g/mol = 215.8 which matches the 216 water molecules found in the standard periodic box. The solvent molecules should be oriented randomly in the box and the entire system equilibrated. You can do this by running Molecular Dynamics on the system. You can then add a solute molecule to the system; you will probably want to delete one or more solvent molecules to make space for it, as HyperChem does with its standard water boxes. An equilibrated box of solvent molecules can be saved as a file, and re-used as you wish. > also what should I do exactly for retrieving information from the www page > of hyperchem?: I think I read that there is a macro for rachamandran plot > available there. Assuming that your WWW browser is set up properly (and that is something for your local network manager to set up), all that you need to do is to go to our site (http://www.hyper.com) and go to the Software Archive. That macro, and a variety of other utilities, are there. Regards, Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" to hyperchem-request@hyper.com WWW: http://www.hyper.com/