From: Sopek@Hyper.Com - using Network [sopek@hyper.com] Sent: Monday, June 01, 1998 1:21 AM To: hyperchem@hyper.com Subject: cyclodextrines Hello ! I promised to post some cyclodextrines to this list. However, years passed when I was involved with these molecules. I dug through all of my old CDs and I found only two HIN molecules and one pretty (I hope) animation. Everything is on my WEB server in Poland: http://www.mako.com.pl/hyper/support/dextrines.htm Feel free for downloading from this page. I'm sorry for rising bigger expectations ! Cheers, Mirek ---- Dr Miroslaw Sopek President and CEO HyperCube Europe, Inc. http://www.hyper.mm.com.pl ---------------------------------------- sopek@hyper.com ------------- From: support [support@hyper.com] Sent: Tuesday, June 02, 1998 3:43 PM To: hyperchem@hyper.com Subject: Ordering the book "Laboratory Experiments..." Dear HyperChemists, There have been many order requests following the recent announcement of the availability of the book "Laboratory Experiments Using HyperChem" by Mary L. Caffery, Paul A. Dobosh, and Diane M. Richardson. The book is priced at $24.95 plus shipping. Please send your order requests to sales@hyper.com or call our sales desk at (352) 371-7744. Regards, Rajiv Rajiv Bendale, 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: Brian Williams [williams@bucknell.edu] Sent: Tuesday, June 02, 1998 7:12 AM To: hyperchem@hyper.com Subject: Lab Experiments Book I would very much like to order a copy of the new textbook on lab exercises using HyperChem. I have started using HyperChem this past spring in two upper level courses, and the students seem to like it very much-in fact, they are now starting to use it to gain insight into other things they are doing, which was my goal. I think this is a wise move on the part of HyperChem, given the extensive educational material available for your competitor Spartan. Do I need to send you a check or cash now, or will I be billed? Brian Williams Department of Chemistry Bucknell University Lewisburg, PA 17837 (717)-524-1143 Fax: (717)-524-1739 From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Tuesday, June 02, 1998 8:30 AM To: HyperChem User's List Subject: Water periodic box Dear All, Periodic box gives an option for setting "Minimum distance between solvent and solute atoms". It shouldn't exceed a half of the largest box dimension (a smallest box including molecule), according to Hyper's RM (p. 222). Does it regard the distance between centers of atoms, or is it thought to be between the Van der Waals surfaces? When displaying such minimal box enclosing molecule, the edges of the box are located at the centers of atoms. Therefore, it is not clear for me whether HyperChem takes into account Van der Waals radii when dealing with interatomic distances. Any ideas are kindly welcomed. Also I'd like to report one more bug: Any SCRIPT execution stops when I'm trying to move a script message box upon the screen. I mean, when I do not touch the script message box then my calculations are running correctly but when I put my mouse pointer on this small box just for moving it the program frozes and no progress in calculations is noticed. Usually the script message box appears in the right-bottom corner of the screen and it so that the button "Abort" is invisible (located outside the screen viewing range). So, for me it's impossible to make any movement of this message box. I am using Windows NT 256 MB RAM with Diamond Fire GL 1000 Pro 8MB. Have you ever noticed such situation? Best regards, Maciej Turowski, Ph.D Kyoto Institute of Technology Japan From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Wednesday, June 03, 1998 3:27 AM To: Ulrich Harttig Cc: HyperChem User's List; Rajiv Bendale Subject: Re: Water periodic box Thank you very much for your message. We have asked for an upgrade of HyperChem Suite 5.1 (containing ChemPlus upgraded, I hope) but our purchasing company (Kubota, Inc., Japan) claims that it won't be available in the nearest future. Thus, we must wait patiently and try to sustain all HyperChem bugs found. I am asking, on that occasion, the people from Hypercube: when do you expect HyperChem Suite 5.1 upgrade to be released and available in Japan? Yours sincerely, Maciej Turowski, Ph. D Kyoto Institute of Technology Japan Ulrich Harttig wrote: > > Also I'd like to report one more bug: Any SCRIPT execution stops when > > I'm trying to move a script message box upon the screen. I mean, whenI > > do not touch the script message box then my calculations are running > > correctly but when I put my mouse pointer on this small box just for > > moving it the program frozes and no progress in calculations is > > noticed. Usually the script message box appears in the right-bottom > > corner of the screen and it so that the button "Abort" is invisible > > (located outside the screen viewing range). So, for me it's impossible > > to make any movement of this message box. I am using Windows NT 256 MB > > RAM with Diamond Fire GL 1000 Pro 8MB. > > In my experience, Hyperchem 5.02 shows that behavior not only with the > script window but also when other windows are open. For instance when > using Molecular dynamics & plotting energy terms, moving the window > obviously disrupts the communication between Hyperchem and the > calculation backend. This is the reason why there is no update in HC > status line while the backend is still happily calculating. > Unless you want to upgrade to HC 5.1 which apparently (as far as I > tested) doesn't has any of these problems, you will have to avoid moving > these windows. > I am using HC 5.02 & HC 5.1 under NT 4.0 + SP3, PPro 200Mhz 64 MB, > Matrox Mystique 4MB > > Best regards > Ulrich > -- > Ulrich Harttig, Ph.D. > Department of Environmental and Molecular Toxicology > Oregon State University, 100 Wiegand Hall, Corvallis, OR, 97331 > Phone: 541-737-6514 Fax: 541-737-1877 From: Dr. Ulrich Jordis [ujordis@pop.tuwien.ac.at] Sent: Wednesday, June 03, 1998 10:01 AM To: hyperchem@hyper.com Subject: Calixarene Who has modelled some Calixarenes and would share the structure with me? ...................................................................... A.o.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: Michael Lococo [ml938@bard.edu] Sent: Friday, June 05, 1998 6:37 AM To: hyperchem@hyper.com Subject: scripting Greetings, Using visual basic and HyperChem Version 3 scripting, I've written a simple little instruction to run an Amber optimization on a system followed by a CNDO single point, to recalculate charge. That process gets looped until the total energy returned by the CNDO calculation is within 5% of the last value. Everything works fine unless my system includes a periodic box filled with TIP3P water molecules. When I switch to Molecular Mechanics with the "calculation-method MolecularMechanics" statement, I get a dialog box asking "Keep current charges? (Instead of replacing with template values)." My wonderful script becomes patently useless if I need to sit there and answer a dialog box for every cycle... HyperChem seems unwilling accept DDE messages until this message box is answered, so I can't answer the message box from visual basic. I have looked in the documentation and readme.doc for switches to the script statement, but can find none. Furthermore, I've tried putting the obvious switches (Yes, No, Y, N, True, etc.) both before and after "MolecularMechanics" in the script statement. I was met each time with some error indicating that there were too many parameters in the statement. My question is: Is this a known problem, or am I a fool for not being able to do it properly? If it is a known problem, is there a fix in newer versions of hyperchem? The Chem deparment I'm working with plans to purchase version 5 for PC as soon as our budget comes through. In the meantime, is there anything I can do besides hand draw and position my water molecules? Sincerely Mike Lococo Bard College Chemistry From: Wayne B. Bosma [bosma@bradley.edu] Sent: Monday, June 08, 1998 9:15 AM To: hyperchem@hyper.com Subject: Hyperchem MD scripts and energy conservation Hi, We're doing molecular dynamics (AM1 and PM3) on a small molecular cluster using a hyperchem script (HC 5.02). Our intention is to run the first part of the simulation at a constant temperature, and then switch to conventional, constant energy, dynamics. The problem seems to be obtaining a constant-energy simulation. We initially just switched "dynamics-constant-temp" to false after the initial constant temperature part; this resulted in the MD reducing the cluster to an overall minimum. We've recently been running the second part with cooling or heating on, but with a bath relaxation time which is very long compared to the simulation. The result has been a simulation in which: 1)the temperature fluctuates a lot during the "constant temp." part, but stays very constant during the "regular" part, and 2) the energy undergoes a slow dip, and then gradually recovers to its original value during the second ("constant energy") part. Any ideas? Thanks, Wayne Bosma Wayne B. Bosma | Dept. of Chemistry | Assistant Professor | Bradley University | Peoria, IL 61625 | |bosma@bradley.edu | (309) 677-3025 | ________________________|______________________| From: Sopek@Hyper.Com - using Network [sopek@hyper.com] Sent: Monday, June 08, 1998 11:17 AM To: Wayne B. Bosma Cc: hyperchem@hyper.com Subject: Re: Hyperchem MD scripts and energy conservation Hello Wayne, The critical issue about semi-quantum-dynamics (as I call what you do for your cluster) being actually classical dynamics on quantum energy surface - is THE QUALITY OF WAVEFUNCTION. If you have poor wave function - the energy will fluctuate. Unlike in a case of classical force fields we have a situation here, that the forces (gradients) at every point of you system trajectory are computed from wavefunction and do not depend on atoms positions, only. (Ideally they, do, however as we do full SCF at every point again and again - they may show wavefunction related differences !) So before, I would comment about behavior you report - please let me know what is the "Accuracy" parameter that you set in Semiempirical Setup window. In my studies of small water clusters I used 10^-6 -> 10^-5 to achieve good wavefunction. Regards, Mirek ---- Dr Miroslaw Sopek President and CEO HyperCube Europe, Inc. http://www.hyper.mm.com.pl ---------------------------------------- sopek@hyper.com ------------- -----Original Message----- From: Wayne B. Bosma To: hyperchem@hyper.com Date: 8 czerwca 1998 12:43 Subject: Hyperchem MD scripts and energy conservation > >Hi, >We're doing molecular dynamics (AM1 and PM3) on a small molecular >cluster using a hyperchem script (HC 5.02). Our intention is to run the >first part of the simulation at a constant temperature, and then switch >to conventional, constant energy, dynamics. >The problem seems to be obtaining a constant-energy simulation. We >initially just switched "dynamics-constant-temp" to false after the >initial constant temperature part; this resulted in the MD reducing the >cluster to an overall minimum. >We've recently been running the second part with cooling or heating on, >but with a bath relaxation time which is very long compared to the >simulation. >The result has been a simulation in which: >1)the temperature fluctuates a lot during the "constant temp." part, but >stays very constant during the "regular" part, and 2) the energy >undergoes a slow dip, and then gradually recovers to its original value >during the second ("constant energy") part. >Any ideas? >Thanks, > >Wayne Bosma > > >Wayne B. Bosma | Dept. of Chemistry | >Assistant Professor | Bradley University | >Peoria, IL 61625 | >|bosma@bradley.edu | (309) 677-3025 | >________________________|______________________| From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Tuesday, June 09, 1998 9:35 AM To: HyperChem User's List Cc: HyperChem Support Subject: London dispersion forces in HyperChem Dear HyperChemists, In the "Hyper Chem - Computational Chemistry" book (Hypercube, Inc.), in the page 39, you cam find following sentence: "You can use CI calculations to do the following: (...) - Capture the effects of London dispersion forces (...)" My question is: Please, tell(show) me HOW??? How exactly can I do it - i.e. "capture the effects of London forces"??? What steps are involved in the procedure cited above? Indices of HyperChem's manuals mention London dispersion forces only once, the same one I said about. I couldn' find, yet, any description regarding exact procedures. If you ever worked with this problem please let me know. I would be very obliged. This is a very important question for me. With my best regards, Sincerely, -- Maciej Turowski, Ph. D turowski@ipc.kit.ac.jp *************************************** ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Kyoto Institute of Technology Department of Polymer Science and Engineering Kyoto 606-8585, Japan //////////////////////////////////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ "Sic transit gloria Mundi" -Julius Caesar From: Loychik Robert James [loychik@bucknell.edu] Sent: Wednesday, June 10, 1998 8:46 AM To: hyperchem@hyper.com Subject: excited state energy For a particular starting molecular geometry, I am interested in using HyperChem to see how the total energy and the dipole moment change for the first six excited states as one particular torsion angle in the molecule is varied. I have been doing this using CI calculations under ZINDO/S where I have been defining the torsion angles manually and obtaining the energy and dipole moment values from log file outputs. Does anyone know of a convenient way to automate this process, such that HyperChem can automatically step through a series of torsion angles and give me just the energy and dipole moment values for these excited states as a function of the torsion angle? Thanks. Bob Loychik loychik@bucknell.edu From: Arthur Cammers [cammers@caribe.chem.uky.edu] Sent: Wednesday, June 10, 1998 1:12 PM To: hyperchem Subject: minimization protocols Dear Hyperchemists: Most often the goal of molecular modeling is finding global or local minima on the conformational hyperspace. Question: Could one not run calculations to maximize properties on the conformational space. For example one might be interested in a set of conformational microstates with maximum dipole moment for a particular molecule. Is such a calculation available in hyperchem? Arthur Cammers-Goodwin Assistant Professor of Chemistry Department of Chemistry-Organic Division University of Kentucky acgood1@pop.uky.edu http://www.chem.uky.edu/research/cammers/cammerscv.html From: facien03@abello.dic.uchile.cl Sent: Monday, June 15, 1998 1:16 PM To: hyperchem@hyper.com Subject: *.ext files. Dear Hyperchemists: I am looking for a program to read the exportation files *.ext and to transform them in ASCII files. It is important for QSAR studies. I will appreciate if someone can help me. Sincerely, Juan S. Gomez-Jeria Associate Professor University of Chile Faculty of Sciences P.O.Box 653 Santiago, CHILE. Phone: (562)-678-7371 Fax: (562)-271-3888. From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Tuesday, June 16, 1998 5:42 AM To: Dieter Klapstein Cc: HyperChem User's List Subject: Re: HyperChem images Dear Dieter, Thank you for your mail. In fact, the method you have described also failed in my case (256 MB RAM was insufficient memory!). Such strange situation was already explained by Hypercube as HyperChem 5.01 bug and it has been fixed in 5.1 version. I hope this upgrade to be sent to me some day... . To overcome that problem in 5.01, however, one can use RGB bitmap format (in Preference menu). Of course it's not a perfect solution but it works. Best greetings, Maciej Turowski Ph. D Kyoto Institute of Technology Japan Dieter Klapstein wrote: > I just came across your query regarding the problem of > HyperChem images becoming degraded when going through the > clipboard. Here is a somewhat indirect work-around we have > found: in the preferences make sure the image is set up as a > bitmap, then set the destination to a FILE, then under edit > copy the image, which will then be saved as the specified > file, which can then be read with other programs (photoshop, > Word, etc.). While not a direct solution, this method gives > you the hard-earned pictures you need. Hope this helps. > > Dieter Klapstein From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Tuesday, June 16, 1998 6:03 AM To: HyperChem Support Cc: HyperChem User's List Subject: Langevin dynamics problem Dear Support, Dear All, I am trying to simulate solvent environment and perform Langevin Dynamics for the system consisted of two hydrocarbons to examine possible dispersion interactions undergoing in the solvent-like conditions. I thought about using either mm+ or amber forcefield. However, when trying to run the following error messages appear: [for Amber] "Bad coordinates received from calculations" [for mm+] "Error message received from node = 0. Two or more atoms are in the same position" It's not true because I built nonane (C9H20) and ethane (C2H6) and placed them in the distance ca. 10 A (by means of translation tool). After assigning charges (Ext. Hueckel) I optimized structures (dielectric distant dependent = 1) and then tried to perform Langevin dynamics (heat time 1 ps from 0 to 300 K, run time 5 ps, cool time 0.5 ps, time step = 0.0005 ps, temp. step = 5 K), unsuccessfully. Also status line and the plot show following values: Total Energy = 1.#QNAN kcal/mol and T = 1.#QNAN K. Have you got any idea about that? Also I would appreciate any suggestions about what values of friction coefficients should be used to mimic different solvents. For water I used literature value of 84 ps-1. Sincerely yours, Maciej Turowski, Ph. D Kyoto Institute of Technology Japan From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Tuesday, June 16, 1998 5:47 AM To: HyperChem Support Cc: HyperChem User's List Subject: Dispersion forces Dear Support, Dear Hyperchemists, In the "Hyper Chem 5 - Computational Chemistry" book (Hypercube, Inc.), in the page 39, you can find a following sentence: "You can use CI calculations to do the following: (...) · Capture the effects of London dispersion forces (...)" My question is: how exactly can I do it - i.e. "capture the effects of London forces"? What steps are involved in the procedure cited above? Will be grateful for your answer, Best greetings, Maciej Turowski, Ph.D Kyoto Institute of Technology Japan From: Support [support@hyper.com] Sent: Tuesday, June 16, 1998 12:45 PM To: 'Maciej Turowski' Cc: Hyperchem List Server Subject: RE: Dispersion forces Dear Dr. Turowski, The Computational Chemistry manual is a general guide to computational chemistry using HyperChem. Some uses or techniques mentioned in the book may, unfortunately, require the user to refer to literature for more detailed information. In your case perhaps reference to the book "Molecular Quantum Mechanics Part III" by P.W. Atkins, Clarendon Press, Oxford,1970, pages 456-460 may more useful. The dispersion (London) forces mentioned in the computational chemistry manual, are attractive forces between two non polar molecules. They arise from the induced-dipole induced-dipole interaction in such molecules. The dispersion force can be obtained from the perturbation due to the induced dipoles on two molecules some distance apart. Since non-polar molecules have no permanent dipole moment, the first order perturbation contribution is zero. The second order energy correction due to the dipoles is an attractive contribution and has a r**(-6) dependence (an attractive component of the Van der Waals interaction) Thus, the dispersion energy is proportional to the mean square fluctuation in the electron density distribution, and in terms of the polarizabilities: E(2)= -X* alpha(a)*alpha(b)/(R**6) Where alpha(a) is the polarizability of molecule a, and alpha(b) is the polarizability of molecule b. R is their separation, E(2) the dispersion energy (London), and X is a constant. It can also be approximated by: E(2)= Y*{I(a)*I(b)}/{I(a)+I(b)}* alpha(a)*alpha(b)/(R**6), Where Y is a constant and I(a), I(b) are the ionization potentials of species a and b. You can thus compute it in this way or you can "capture the effect of these forces" as follows: Carry out a CI calculation of each of the two molecules A and B separately. The sum of the CI energies of A and B is your reference zero energy. Then carry out a CI calculation on the super molecule A+B where A and B are separated by some distance R. Generate a plot of the CI energy of this super molecule, for a number of values of R. The depth of the potential well with respect to the reference zero energy is your dispersion energy or the 'attractive energy' holding the two non-polar molecules together. Regards, Rajiv Rajiv Bendale, 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 [mailto:turowski@ipc.kit.ac.jp] Sent: Tuesday, June 16, 1998 5:47 AM To: HyperChem Support Cc: HyperChem User's List Subject: Dispersion forces Dear Support, Dear Hyperchemists, In the "Hyper Chem 5 - Computational Chemistry" book (Hypercube, Inc.), in the page 39, you can find a following sentence: "You can use CI calculations to do the following: (...) · Capture the effects of London dispersion forces (...)" My question is: how exactly can I do it - i.e. "capture the effects of London forces"? What steps are involved in the procedure cited above? Will be grateful for your answer, Best greetings, Maciej Turowski, Ph.D Kyoto Institute of Technology Japan From: Gerardo Burton [burton@qo.fcen.uba.ar] Sent: Tuesday, June 16, 1998 3:53 PM To: Maciej Turowski Cc: Hyperchem Subject: RE: Langevin dynamics problem Hello Maciej I had a similar problem some time ago but with normal dynamics and the molecular mechanics force fields in one of our PCs. I found out that when running the upgrade patch from version 5.0 to 5.02 in that machine some of the backend files (mmplus.exe, newton.exe, ndo.exe, gauss.exe, etc) had not been updated by the patch (I don't know why), this doesn't seem to affect most of the calculations except dynamics with the mm+ and amber force fields (no problem running dynamics with the semiempirical methods), so check those files, they should be dated september 1997 not 1996 (mmplus.exe should be 367 Kb, instead of 232 Kb, etc...) Good luck Gerardo Prof. Gerardo Burton Departamento de Quimica Organica Facultad de Ciencias Exactas y Naturales Universidad de Buenos Aires e-mail: burton@qo.fcen.uba.ar Tel/Fax: 54 1 788-6915 -----Original Message----- De: Maciej Turowski Para: HyperChem Support CC: HyperChem User's List Fecha: martes 16 de junio de 1998 15:54 Asunto: Langevin dynamics problem >Dear Support, Dear All, >I am trying to simulate solvent environment and perform Langevin >Dynamics for the system consisted of two hydrocarbons to examine >possible dispersion interactions undergoing in the solvent-like >conditions. I thought about using either mm+ or amber forcefield. >However, when trying to run the following error messages appear: >[for Amber] "Bad coordinates received from calculations" >[for mm+] "Error message received from node = 0. Two or more atoms are >in the same position" >It's not true because I built nonane (C9H20) and ethane (C2H6) and >placed them in the distance ca. 10 A (by means of translation tool). >After assigning charges (Ext. Hueckel) I optimized structures >(dielectric distant dependent = 1) and then tried to perform Langevin >dynamics (heat time 1 ps from 0 to 300 K, run time 5 ps, cool time 0.5 >ps, time step = 0.0005 ps, temp. step = 5 K), unsuccessfully. >Also status line and the plot show following values: Total Energy = >1.#QNAN kcal/mol and T = 1.#QNAN K. >Have you got any idea about that? >Also I would appreciate any suggestions about what values of friction >coefficients should be used to mimic different solvents. For water I >used literature value of 84 ps-1. >Sincerely yours, > >Maciej Turowski, Ph. D >Kyoto Institute of Technology >Japan > From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Wednesday, June 17, 1998 5:48 AM To: Support Cc: HyperChem User's List; Mirek Sopek Subject: Re: Langevin dynamics problem Dear Hyperchemists, Dear Dr Bandale, Dear Dr Sopek, In fact, I am using HyperChem 5.02. I have replaced mmplus.exe and newton.exe by those from Hypercube's site. Now, when I'm trying to perform Langevin dynamics, the following message appears: "Floating point error encountered. May have been caused by an ill-conditioned molecular system". After pressing "Ignore" I got the message known already: "Two or more atoms are in the same position". So I can't perform Langevin althoug the same system has been found appropriate to run Molecular Dynamics. With this last one there is no problem. ***************** One additional thing is that when finally I could perform amber molecular dynamics in the water box in my hin file there were errors that made my file impossible to open by HC, something like that: forcefield amber sys 1.#QNAN view 40 0.043649 88.3 48.3 1 0 0 0 6.123032e-017 1 0 -1 6.123032e-017 0.098214 0.19297 -88.276 box 31.35789 31.35789 31.35789 seed 0 dynamics C:\Hyper5\Program\Dane\H_D\C18\proba1.snp mol 1 atom 1 - C CT - -0.08706617 3.900116 -0.05822448 -3.076487 4 2 s 8 s 9 s 10 s vel 1 1.#QNAN 1.#QNAN 1.#QNAN atom 2 - C CT - -0.02268124 2.412337 0.2998644 -3.084368 4 1 s 3 s 11 s 12 s vel 2 1.#QNAN 1.#QNAN 1.#QNAN etc... Also *.csv file was like that: time, EKIN, EPOT,D EKIN,D EPOT 0,1.#QNAN,60.7183,0,0.578821 So I couldn't neither playback MD nor open the molecule that was used for MD. I will attach the exemplary files in the next messages to you (Dr Bandale and Dr Sopek). Those will be the file I am trying to perform Langevin Dynamics on and the bad hin file from MD. Greetings, Maciej Turowski, Ph. D Kyoto Institute of Technology Japan Support wrote: > Dear Dr. Turowski, > > It appears that you are using HyperChem 5.0 , 5.01 or 5.02 versions. Do > you get the error when the cooling cycle begins? If so, you need to > apply a patch available on our web site at > http://www.hyper.com/Support/MDpatch.html. This is an old bug fix and > has been fixed a while ago. Please let me know if this is the case with > your version of HyperChem. > > Regards, > Rajiv > --- > Rajiv Bendale, 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: Mirek Sopek [sopek@hyper.com] Sent: Tuesday, June 16, 1998 11:19 PM To: Maciej Turowski Cc: hyperchem@hyper.com Subject: Re: Langevin dynamics problem Dear Dr Turowski, I ran your alkanes.hin on the HyperChem 5.01 installation. I confirm that there was the error that you report, however, as Dr Bandale suspected the error was only at the beginning of cooldown phase of Langevin Dynamics and has been fixed quite a time ago. The patch is available. Thank you for sending files - from your initial message it wasn't clear when exactly you receive the error - and I was "almost" sure that "normal" Langevin was working fine as long as you did not use simulated annealing technique. Since the patch and in the HyperChem 5.1 family this technique works fine, too. Mirek ---- Dr Miroslaw Sopek President and CEO HyperCube Europe, Inc. http://www.hyper.mm.com.pl ---------------------------------------- sopek@hyper.com ------------- From: Maciej Turowski [turowski@ipc.kit.ac.jp] Sent: Wednesday, June 17, 1998 9:41 AM To: Mirek Sopek Cc: hyperchem@hyper.com; HyperChem Support Subject: Re: Langevin dynamics problem Dear Dr Sopek, I am very affraid to say but the error appeared just at the beginning of the process i.e. not cooling but heating period. So it was just the first step of Langevin dynamics - from 0 ps to 0.005 ps. Moreover, new file downloaded from Hypercube did not help. The same problem appeared, as I mentioned in my last e-mail to you and to Dr Bendale. Sorry about that. Yours sincerely, Maciej Turowski KIT Japan P.S. I am enclosing my e-mail about it: Dear Hyperchemists, Dear Dr Bandale, Dear Dr Sopek, In fact, I am using HyperChem 5.02. I have replaced mmplus.exe and newton.exe by those from Hypercube's site. Now, when I'm trying to perform Langevin dynamics, the following message appears (at the beginning of a heat time): "Floating point error encountered. May have been caused by an ill-conditioned molecular system". After pressing "Ignore" I got the message known already: "Two or more atoms are in the same position". So I can't perform Langevin althoug the same system has been found appropriate to run Molecular Dynamics. With this last one there is no problem. ***************** One additional thing is that when finally I could perform amber molecular dynamics in the water box in my hin file there were errors that made my file impossible to open by HC, something like that: forcefield amber sys 1.#QNAN view 40 0.043649 88.3 48.3 1 0 0 0 6.123032e-017 1 0 -1 6.123032e-017 0.098214 0.19297 -88.276 box 31.35789 31.35789 31.35789 seed 0 dynamics C:\Hyper5\Program\Dane\H_D\C18\proba1.snp mol 1 atom 1 - C CT - -0.08706617 3.900116 -0.05822448 -3.076487 4 2 s 8 s 9 s 10 s vel 1 1.#QNAN 1.#QNAN 1.#QNAN atom 2 - C CT - -0.02268124 2.412337 0.2998644 -3.084368 4 1 s 3 s 11 s 12 s vel 2 1.#QNAN 1.#QNAN 1.#QNAN etc... Also *.csv file was like that: time, EKIN, EPOT,D EKIN,D EPOT 0,1.#QNAN,60.7183,0,0.578821 So I couldn't neither playback MD nor open the molecule that was used for MD. I will attach the exemplary files in the next messages to you (Dr Bandale and Dr Sopek). Those will be the file I am trying to perform Langevin Dynamics on and the bad hin file from MD. Greetings, Maciej Turowski, Ph. D Kyoto Institute of Technology Japan Mirek Sopek wrote: > Dear Dr Turowski, > > I ran your alkanes.hin on the HyperChem 5.01 installation. > I confirm that there was the error that you report, however, as Dr Bandale > suspected the error was only at the beginning of cooldown phase of Langevin > Dynamics and has been fixed quite a time ago. The patch is available. > > Thank you for sending files - from your initial message it wasn't clear when > exactly > you receive the error - and I was "almost" sure that "normal" Langevin was > working fine > as long as you did not use simulated annealing technique. > Since the patch and in the HyperChem 5.1 family this technique works fine, > too. > > Mirek > ---- > Dr Miroslaw Sopek > President and CEO > HyperCube Europe, Inc. http://www.hyper.mm.com.pl > ---------------------------------------- sopek@hyper.com ------------- From: Mirek Sopek [sopek@hyper.com] Sent: Wednesday, June 17, 1998 9:50 AM To: Maciej Turowski Cc: hyperchem@hyper.com Subject: Re: Langevin dynamics problem Dear Dr Turowski, I do not quite understand what is going on with this error. I used my installation of regular HCH 5.01 (without patch) and I did not have any error on YOUR HIN file unless I used explicit cooling phase in the Langevin. Could you create a HCL (*.scr) script that drives your Dynamics completely, so we could eliminate some settings that were made "by hand" ? Write everything there what you do for the LD, time steps, temperatures, etc. Now - when you have it, and it still crashes in the first steps - could you send it to us ? Please do this on a fresh installation with appropriate patch installed ! Mirek Dr Miroslaw Sopek President and CEO HyperCube Europe, Inc. http://www.hyper.mm.com.pl ---------------------------------------- sopek@hyper.com ------------- From: Ed Bertaccini, MD [bert@vapa.icon.palo-alto.med.va.gov] Sent: Monday, June 29, 1998 11:38 AM To: hyperchem@hyper.com Subject: Amber5 Dear Hyperchem User Group, Has anyone had any success importing the new Amber 5.0 parameter sets into Hyperchem? Thanks, -------signature file attached to all documents------- Ed Bertaccini, MD Assistant Professor of Anesthesia, Stanford University School of Medicine Staff Anesthesiologist and Intensivist, Palo Alto VA Health Care System VA office phone: 415-493-5000 ext. 65180 VA office fax: 415-852-3423 From: UH [harttigu@ava.bcc.orst.edu] Sent: Monday, June 29, 1998 9:15 PM To: Ed Bertaccini, MD Cc: hyperchem@hyper.com Subject: Re: Amber5 I had exactly the same question about a year ago. I got no reply so I started teh implementation on my own. This work was finished earlier this spring. Here is the current status of this project: I implemented the AMBER94 FF parameters (parm94.dat/parm96.dat) for HyperChem 5.02/5.1 (HC), including appropriate new atom types, parameters, type rules and AA/DNA templates + charges. Single-point calculations of 5 protein structures were used to compare this implementation with other program packages using the AMBER FF : The prgrams used were: TINKER 3.6, MMTK 1.1b (calculations done by Konrad Hinsen) and AMBER 5.0. The comparison showed that the Amber94 implementation for HC produced wrong dihedral energy terms while the other energy terms were mostly close to the values from the other programs. Upon further inspection of this problem, I discovered that the error was caused by torsion angle parameters with a periodicity of N=4. Although there are only three terms in the AMBER94 parameter set which generate this error (there are none in the old AMBER91 parameter set), unfortunately one of them is the torsion parameter for N-CT-C-N in the peptide bond. The most recent modification of the AMBER parameter set (parm96.dat) eliminates the N=4 term from the N-CT-C-N torsion parameter but the remaining parameters (CT-CT-C-N, CT-CT-N-C) still cause wrong dihedral energies in HC. I contacted HyperCube in March '98 about this problem, but have not heard anything from them with respect to this calculation problem. I am pretty confident that my implementation is not at fault but that a bug in HC prevents currently the use of the most recent AMBER FF parameters. If you have further questions please feel free to contact me. For a detailed look at the comparative calculations please, check the following URL: http://128.193.211.62/wwwdata/amber-comparison.htm Regards Ulrich Harttig, Ph.D. Department of Environmental and Molecular Toxicology Oregon State University, 100 Wiegand Hall, Corvallis, OR, 97331 Phone: 541-737-6514 Fax: 541-737-1877 From: Ed Bertaccini, MD [bert@vapa.icon.palo-alto.med.va.gov] Sent: Tuesday, June 30, 1998 11:52 AM To: UH Cc: hyperchem@hyper.com Subject: Re: Amber5 Dear Dr. Harttig, Thank you for such in depth information regarding the implementation of parm94 and parm96 parameters in Hyperchem 5.1. I am disappointed that Hyperchem has not been more responsive. Perhaps this discussion will spur on some interest. I feel that this is a very important issue because the new parameters have been derived in what is probably a more valid way (i.e. the RESP technique for partial atomic charges) than previously. These need to be implemented in the current version of Hyperchem so as to lend further credence to what many UNIX exclusivists consider "just a PC program." It is very obvious that the PC workstation realm is becoming a very valid platform for highly technical computing, especially in light of the recent plan set forth by SGI. Hyperchem is fantastic software and its interface in particular brings readily complex capabilities to those of us who have neither the time nor the patience to prepare a myriad of input files for more command prompt-based software packages. I look forward to further input on this subject, -------signature file attached to all documents------- Ed Bertaccini, MD Assistant Professor of Anesthesia, Stanford University School of Medicine Staff Anesthesiologist and Intensivist, Palo Alto VA Health Care System VA office phone: 415-493-5000 ext. 65180 VA office fax: 415-852-3423