From ostlund@hyper.hyper.com Tue Oct 4 01:20:06 1994 Subject: MM+ To: hyper!hyperchem@uunet.ca Date: Mon, 3 Oct 1994 21:20:06 -0400 Cc: hyper!ostlund@uunet.ca (Neil S. Ostlund) Morten Langgaard recently raised a question about the philosophy behind MM+. It seems like maybe I should respond to this since it does in a sense involve why rather than just what. I don't want to give the impression that we are dogmatic at Hypercube - we like to give people what they want - but we usually have reasons for why we do what we do. I guess the short answer is that we are not enamoured of the MMP philosophy requiring quantum mechanical computation in molecular mechanics approaches and that we think MM+ may be a better pure mechanics approach. If we felt MMP was essential we would have implemented it or if we think users now are starting to demand it we will seriously look at putting it in. Most of you are aware of the general shortcomings of MM2, MM3, Amber and most of the molecular mechanics techniques that base their parameters solely on the atom-type of the relevant atoms. This leads to equal bond lengths in butadiene, biphenyl, etc. The solution to this problem is to include some consideration of the actual type of bond, bond-order, etc. in defining the bond parameters. One way to do this is to do a "pi" kind of calculation like MMP. As we all know, however, their is no rigorous definition of a pi system in the general 3D world of real molecules. Another very general approach would be to always do an extended Huckel calculation on the molecule to define the bond-orders and hence molecular mechanics parameters. The problem with this is, of course, one is trying to do a rapid calculation in the molecular mechanics style but one now has the problem that one is doing quantum mechanical calculations on the whole molecule rather than (rapid?) molecular mechanics. The approach we like, which derives from W.Goddard et al., is the Dreiding or Universal Force Field (UFF) approach which argues for a specific bond-order from the type of bond, hybridization of the relevant involved atoms, and other chemistry notions. MM+ incorporates these ideas into a general scheme for deriving molecular mechanics parameters totally independent of MM2. MM+ is really two force fields - our own version of the Dreiding, UFF approach plus the MM2 force field. The way it works is that when MM2 parameters are available it uses them. When they are not it uses the Dreiding, UFF parameters to compliment the MM2 parameters as fully described in the manuals. This hopefully means one can have one's cake and eat it also. For example, one can trivially tell HyperChem that no MM2 parameters are available by just setting the atom-types to ** (globally or for any selected subset of atoms). Thus MM+ gives a standard MM2 answer of equal bond lengths for butadiene in the normal situation (when atom-types have been assigned) but if one quickly just sets all the atom types to **, one gets the MM+ (Dreding, UFF) correct answer that the central bond is much longer than the terminal bonds. All this without having to do relatively ambiguous quantum mechanical pi calculations. In addition, it means that no calculation fails for the lack of a molecular mechanics parameter. Obviously, there are many other facets on this general topic of what are good, bad, and mediocre (the right and wrong?) molecular mechanics methods and truth is very rarely obtained but I hope this explains a little bit about why we don't currently do MMP calculations. The general question of introducing ideas from both molecular mechanics and quantum mechanics into the same calculation is one in which we have a great deal of interest, as evidenced by HyperNMR and its "typed neglect of differential overlap" method where we introduce the idea of atom-types into quantum mechanical calculations. Neil Ostlund ________________________________________________________________________ From BLACKMAN@alkali.otago.ac.nz Wed Oct 5 15:18:46 1994 From: "Allan Blackman" To: hyperchem@HYPER.COM Date: Wed, 5 Oct 1994 11:18:46 -0400 Subject: Z-matrix files? Reply-To: alkali.otago.ac.nz!blackman@hyper.hyper.com Hi - please excuse me if this has been asked before but... how do get HyperChem (release 3) to save files in z-matrix form suitable for imputting into MOPAC? I've tried simply saving the file as *.zmt, but this just gives a .hin format under another name! Thanks for your help... Cheers Allan ********************************************************************** Allan Blackman, Department of Chemistry, University of Otago, Dunedin, New Zealand. e-mail blackman@alkali.otago.ac.nz Marge on innocence/guilt "..as my uncle used to say, shoot them all and let God decide...." ********************************************************************** ________________________________________________________________________ From DBellnie@hs1.buffalo.edu Wed Oct 5 16:22:00 1994 From: "Bellnier, David" To: hyperchem@HYPER.COM Date: Wed, 5 Oct 1994 12:22:00 -0400 Is there a way to calculate, using Hyperchem, the *relative* amphiphilicities of a series of simple (porphyrin-based) molecules? These molecules are amphiphilic because on one side of the molecule is a carboxylic acid function (hydrophilic) and on the opposite side is an aliphatic chain (hydrophobic) which varies in length between the different compounds in the series. Is there a measurable "amphiphilic moment" between these two groups? ________________________________________________________________________ From lmarkham@darkwing.uoregon.edu Wed Oct 5 22:18:35 1994 Date: Wed, 5 Oct 1994 18:18:35 -0400 From: Laura Markham To: hyperchem@HYPER.COM Subject: problems with exporting I am using hyperchem to visualize my calcuated normal modes. I need to save these figures to a format readable by Wordperfect or WP presentations. My problem is this.... The IR vectors do not rescale when I scale to fit the molecule. The vectors are then out of proportion to the molecule. Although I can print out the molecule with vectors (inc the off screen stuff) directly to the printer I have been unable to save it to a transferable file. (It cuts all of the off screen stuff off) Does anyone know of a reasonable way to do this? Thanks Laura Markham Chemistry Dept. Univ of OR Eugene, OR 97403 lmarkham@darkwing.uoregon.edu ________________________________________________________________________ From hurst@hyper.hyper.com Wed Oct 5 23:53:11 1994 From: hurst@hyper.hyper.com (Graham Hurst) To: Laura Markham Subject: Re: problems with exporting Cc: hyperchem@HYPER.COM Laura Markham writes: > I am using hyperchem to visualize my calcuated normal modes. I need to > save these figures to a format readable by Wordperfect or WP presentations. > My problem is this.... The IR vectors do not rescale when I scale to fit > the molecule. The vectors are then out of proportion to the molecule. > Although I can print out the molecule with vectors (inc the off screen stuff) > directly to the printer I have been unable to save it to a transferable file. > (It cuts all of the off screen stuff off) Thanks for pointing this out! I think they should scale with the image zoom factor and ideally have the vector lengths scalable by the user, so I'll add this to the bug list. > Does anyone know of a reasonable way to do this? I'm not sure whether it's reasonable, but there is a workaround. It turns out that the vectors do scale with perspective, so if you toggle on perspective you can Z translate the view to make the molecule and the vectors smaller, and them adjust the zoom to make the molecule (but not the vectors) larger. The result will be relatively smaller vectors. Of course perspective will be on, but this will be less apparent the more distant you make the view. (You will probably also have to adjust the Z clipping back plane to still see the molecule.) Once you've got it looking okay on screen you can use the usual Edit/Copy Image to get a metafile and/or bitmap in the clipboard and/or file(s) that you should be able to use in WordPerfect. Hope this helps, Graham ------------ Graham Hurst (hurst@hyper.com) Hypercube Inc, 7-419 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 ________________________________________________________________________ From ELEC!UNDERHILLR%MECH_STAFF@BANEE.rmc.ca Thu Oct 6 12:44:04 1994 Received: by BANEE.rmc.ca with VINES ; Thu, 6 Oct 94 10:02:01 EDT Date: Thu, 6 Oct 1994 08:44:04 -0400 From: BANEE.rmc.ca!UNDERHILLR%MECH_STAFF%ELEC@hyper.hyper.com Subject: amphiphobicity To: hyperchem@HYPER.COM Is there a way to calculate, using Hyperchem, the *relative* amphiphilicities of a series of simple (porphyrin-based) molecules? These molecules are amphiphilic because on one side of the molecule is a carboxylic acid function (hydrophilic) and on the opposite side is an aliphatic chain (hydrophobic) which varies in length between the different compounds in the series. Is there a measurable "amphiphilic moment" between these two groups? This is an interesting question and one that I would love an answer to as well. The work of Politzer et al. suggests that extremely good measures of hydrophobicity etc. can be obtained by measuring the deviation of the ESP on the Van der Waals surface of the molecule. A number of other users have also asked about calculating ESP based charges, such as ChelpG, which also require knowing the ESP on the Van der Waal surface. At the moment there appears to be no way of doing this within Hyperchem. Perhaps if enough of us ask for it Graham and company will oblige. In your particular case the best I could suggest is to measure the charge moment of the molecule based on a Mulliken analysis and maybe compare this with the first moment of inertia of the molecule. These should be easy to calculate. An empirical relationship between amphiphilicity determined experimentally (try looking at data on Langmuir films and the solubility of linear chain carboxylic acids and alcohols) and the various moments or combinations thereof might be established. A fair amount of work but it would make a good paper. ________________________________________________________________________ From lmarkham@darkwing.uoregon.edu Wed Oct 12 03:29:59 1994 Date: Tue, 11 Oct 1994 23:29:59 -0400 From: Laura Markham To: hyperchem@HYPER.COM Subject: exporting to wordperfect Hi Again! I'm trying to put my figures (which print very nicely directly from Hyperchem) into wordperfect documents. I have done this by copying to the clipboard and then pasting. The figures form regular hyperchem and presentations look ok in wordperfect on the screen. However, when I print from there I have problems. The lt blue and black spheres (which were distinguishable printed from hyperchem) are not distinguishable. Also, The labels on the atoms come out either miniscully small or very large depending on which printer I use. On the screen everything looks ok. I am using windows 3.1, wpwin6.0a, and either an HPIIIsi or HP4-PS printer. Does anyone out there have experience getting there figures into wordprocessing documents successfully. Help please! Aside, I've been getting messages from postmasters telling me my message was undeliverable, although it is forwarded to me. Is this just form someone who no longer has an account. Is there a way to stop it? THANK YOU Laura Markham Chemistry Dept. University of Oregon ________________________________________________________________________ From DESJARDI.S@fs.sciences.WLU.EDU Thu Oct 13 17:02:21 1994 To: hyperchem@HYPER.COM From: "Dr. Steven Desjardins" Organization: Washington & Lee University Date: Thu, 13 Oct 1994 13:02:21 -0400 Subject: boat-chair transition I was wondering if there was a way to move one end of a cyclohexane molecule to go from chair to boat without using a reflection plane. I'm trying to illustrate how the hydrogens switch from equatorial to axial, but if I label the hydrogens and reflect, the axial ones are still axial. Thanks. Steven G. Desjardins Department of Chemistry Washington and Lee University Lexington VA 24450 (703) 463-8873 ________________________________________________________________________ From g112976@iris.ufscar.br Thu Oct 13 16:29:28 1994 Date: Thu, 13 Oct 1994 12:29:28 -0400 From: iris.ufscar.br!g112976@hyper.hyper.com (Augusto C. C. Neto) Subject: Garnet Structures To: hyperchem@HYPER.COM Federal University at Sao Carlos, october 13, 1994 Hi Netters !! I'm begining to work with garnet structures, I would like to find some coor- dinates or inputs for Mopac or Hyperchem ... Thank you ... Augusto Cesar ________________________________________________________________________ From hurst@hyper.hyper.com Thu Oct 13 18:38:46 1994 From: hurst@hyper.hyper.com (Graham Hurst) To: "Dr. Steven Desjardins" Subject: Re: boat-chair transition Cc: hyperchem@HYPER.COM Dr. Steven Desjardins writes: > I was wondering if there was a way to move one end of a cyclohexane > molecule to go from chair to boat without using a reflection plane. I'm > trying to illustrate how the hydrogens switch from equatorial to axial, > but if I label the hydrogens and reflect, the axial ones are still axial. > Thanks. One way is to do a restrained optimization. Try an AMBER optimization of the HIN file below. I set it up by drawing and building (chair) cyclohexane, selecting and naming two opposite dihedral angles, restraining them to their chair values of 60 and -60 with force constants of 100, then reflecting one "end" to make boat. ---------------cut here------------------------------- ;Boat cyclohexane with restraints to make it chair forcefield amber sys 0 view 40 0.2409292 55 15 0.02743927 0.221568 0.9747588 -0.9882096 0.1529495 -0.006948362 -0.1506284 -0.9630753 0.2231523 0.6341427 -0.611938 -54.79973 mol 1 atom 1 - C CT - 0 -1.435846 -0.2178103 -1.045265 4 2 s 6 s 7 s 8 s atom 2 - C CT - 0 -1.435844 1.32219 -1.045268 4 1 s 3 s 9 s 10 s atom 3 - C CT - 0 0.01608519 1.835513 -1.04527 4 2 s 4 s 11 s 12 s atom 4 - C CT - 0 0.7420397 1.322178 0.2121382 4 3 s 5 s 13 s 14 s atom 5 - C CT - 0 0.7420379 -0.2177991 0.2121418 4 4 s 6 s 15 s 16 s atom 6 - C CT - 0 0.01608193 -0.7311381 -1.045265 4 5 s 1 s 17 s 18 s atom 7 - H HC - 0 -1.949669 -0.5811397 -0.1552774 1 1 s atom 8 - H HC - 0 -1.94968 -0.5811477 -1.935242 1 1 s atom 9 - H HC - 0 -1.949666 1.685524 -0.1552823 1 2 s atom 10 - H HC - 0 -1.949678 1.685524 -1.935247 1 2 s atom 11 - H HC - 0 0.01608808 2.925514 -1.045266 1 3 s atom 12 - H HC - 0 0.5299159 1.472179 -1.935251 1 3 s atom 13 - H HC - 0 0.2281969 1.685508 1.102114 1 4 s atom 14 - H HC - 0 1.769701 1.685512 0.2121463 1 4 s atom 15 - H HC - 0 1.7697 -0.5811311 0.2121434 1 5 s atom 16 - H HC - 0 0.228202 -0.5811278 1.102122 1 5 s atom 17 - H HC - 0 0.5299107 -0.3677928 -1.935242 1 6 s atom 18 - H HC - 0 0.01609051 -1.821138 -1.04527 1 6 s endmol 1 selection tor1 4 selectrestraint 2 100.000000 60.000000 0.000000 0.000000 0.000000 selectatom 3 molecule 1 selectatom 4 molecule 1 selectatom 5 molecule 1 selectatom 6 molecule 1 endselection tor1 selection tor2 4 selectrestraint 2 100.000000 -60.000000 0.000000 0.000000 0.000000 selectatom 3 molecule 1 selectatom 2 molecule 1 selectatom 1 molecule 1 selectatom 6 molecule 1 endselection tor2 ---------------cut here------------------------------- Hope this helps, Graham ------------ Graham Hurst (hurst@hyper.com) Hypercube Inc, 7-419 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 ________________________________________________________________________ From louey@SHU.sacredheart.edu Fri Oct 14 00:33:31 1994 Date: Thu, 13 Oct 1994 20:33:31 -0400 From: SHU.sacredheart.edu!louey@hyper.hyper.com (louey) To: hyperchem@HYPER.COM Subject: sophomore organic applications? Hi Netters! I am interested in what kinds of ways hyperchem could be used in teaching sophomore organic chemistry. I am knowing what has been done. THank you, Jim Louey Department of Chemistry Sacred Heart University Fairfield, CT 06432-1000 (203) 371-7793 ________________________________________________________________________ From simpson@biobase.aau.dk Fri Oct 14 13:43:19 1994 Subject: seq ed To: hyperchem@HYPER.COM (hyperchem) Date: Fri, 14 Oct 1994 09:43:19 -0400 From: David Simpson Hi Netters, I have a problems with putting an oligopeptide sequence into chemplus from sequence editor if the secondary structure is given as entirely random. Any suggestions? Regards, David Simpson Carlsberg Laboratory. Gl. Carlsbergvej 10, DK 2500 Valby. Denmark. ________________________________________________________________________ From hurst@hyper.hyper.com Fri Oct 14 17:01:49 1994 From: hurst@hyper.hyper.com (Graham Hurst) To: David Simpson Subject: Re: seq ed Cc: hyperchem@HYPER.COM David Simpson writes: > I have a problems with putting an oligopeptide sequence into > chemplus from sequence editor if the secondary structure is given as > entirely random. Any suggestions? > > Regards, > David Simpson > Carlsberg Laboratory. > Gl. Carlsbergvej 10, > DK 2500 Valby. Denmark. What is the nature of the problem? "Random" for sequence editor is just that - a random number generator is used to generate phi and psi angles. This is admittedly rather useless and may give poor or even ridiculous structures - is that the problem? Cheers, Graham ------------ Graham Hurst (hurst@hyper.com) Hypercube Inc, 7-419 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 ________________________________________________________________________ From CCATF@VM.LNCC.BR Tue Oct 18 08:56:40 1994 Date: Tue, 18 Oct 1994 04:56:40 -0400 From: VTBIT.CC.VT.EDU!CCATF%BRLNCC.BITNET@hyper.hyper.com To: hyperchem@HYPER.COM hyperchem: please inform me if chemplus works with release 4 hyperchem 4. what advantages does chemplus have. ________________________________________________________________________ From underhil@hp.rmc.ca Fri Oct 28 12:30:49 1994 Date: Fri, 28 Oct 1994 08:30:49 -0400 To: hyperchem@HYPER.COM From: hp.rmc.ca!underhil@hyper.hyper.com (Ross Underhill) Subject: excel script bug? I was happily playing away using excel to drive a calculation using DDE when I came across a curious(dare I say) bug. I kept getting an error mesage when it tried to execute the lines [delete-atom 10 1] and [delete-atom 9 1]. Both of these lines worked just fine from a script and the sequence [select-atom 9 1][delete-selected-atoms] also worked just fine. In the end I used the latter and happily filled several spreadsheets with data. For the record I was using Excel 5.0 (rather a nice package). I have another slight problem, which I suspect can't be solved but I would happily accept any suggestions. I'm trying to simulate a liquid of dimethyl methyl phosphonate (CH3)PO(OCH3)2 using OPLS united atoms. CH3 | CH3-O-P-O-CH3 " O To develop the extra parameters I need (primarily torsions about the P-O bonds) I'm using PM3 calculations. (PM3 gave the best IR spectrum and compared reasonably well with ab initio calculations). So far I've discovered that I need to scale up the 1-4 electrostatic interactions (no problem) but I also have to scale down the Van der Waals interaction between the C in the OCH3 groups which is too high. I've set the parameters back a bit to that of C ( as opposed to C3) but this is not enough and I'm reluctant to fiddle with these parameters any more. Unfortunately these atoms are 1-5 to each other so I can't use the 1-4 scaling factor. About the only thing left that I can think to do is lengthen the P-O bond to shove the atoms further apart. Not a nice solution. Do you have any suggestions. Do you have any experience in parameterizing the molecular mechanics methods using the semi-empirical ones? If you don't have an easy answer could I ask you to post the second half of this message to the list to see if any other users have some experience with this. Thanks I've now tried to popst this message twice to help@hyper.com and had it returned as address unknown! Dr. Ross Underhill Royal Military College of Canada Kingston, Ontario (613) 541-6000 X6175 ________________________________________________________________________ From underhil@hp.rmc.ca Fri Oct 28 15:09:29 1994 Date: Fri, 28 Oct 1994 11:09:29 -0400 To: hyperchem@HYPER.COM From: hp.rmc.ca!underhil@hyper.hyper.com (Ross Underhill) Subject: Parameterizing OPLS Sorry if a copy of this has already hit the list but.... I have a slight problem, which I suspect can't be solved but I would happily accept any suggestions. I'm trying to simulate a liquid of dimethyl methyl phosphonate (CH3)PO(OCH3)2 using OPLS united atoms. CH3 | CH3-O-P-O-CH3 " O To develop the extra parameters I need (primarily torsions about the P-O bonds) I'm using PM3 calculations. (PM3 gave the best IR spectrum and compared reasonably well with ab initio calculations). So far I've discovered that I need to scale up the 1-4 electrostatic interactions (no problem) but I also have to scale down the Van der Waals interaction between the C in the OCH3 groups which is too high. I've set the parameters back a bit to that of C ( as opposed to C3) but this is not enough and I'm reluctant to fiddle with these parameters any more. Unfortunately these atoms are 1-5 to each other so I can't use the 1-4 scaling factor. About the only thing left that I can think to do is lengthen the P-O bond to shove the atoms further apart. Not a nice solution. Does anyone have any suggestions? Does antone have any experience in parameterizing the molecular mechanics methods using the semi-empirical ones? Thanks Dr. Ross Underhill Royal Military College of Canada Kingston, Ontario (613) 541-6000 X6175