From chemistry-request@ccl.net Mon Jul 20 02:45:43 1992 Date: Mon, 20 Jul 92 00:30:23 CDT From: Ron Elber To: chemistry@ccl.net Status: R Anonuncing the availability of the public domain program MOIL - MOlecular dynamics at ILlinois MOIL is a package of separate programs to simulate heteregenous systems such as proteins. It was developed at University of Illinois at Chicago, Department of Chemistry by Ron Elber, Adrian Roitberg, Gennady Verkhivker, Robert Goldstein, Haiying Li and Carlos Simmerling. Other contributers that helped in testing, developing parameter sets and writing documentation: Danuta Rojewska, Chyung Choi and Alex Ulitsky. This code can be used as yet another molecular dynamics program. For example, the empirical potential used in this program is a combination of a covalent part taken primarily from AMBER (Weiner et al JACS 106,765(1984)) (Improper torsion terms were taken however from CHARMM paper Brooks et al. J.Comp.Chem. 4,187(1983)) and nonbonded parameters taken from OPLS (Jorgensen and Tirado-Rives JACS 110,1657(1988)). This program is doing some standard things, like energy calculation, energy minimization and dynamics. Summary of standard features follows: (i) building energy surface for macromolecules (with existing database for proteins based on AMBER/OPLS force field). (ii) single energy calculation (iii) energy minimization (conjugate gradient and truncated newton-raphson) (iv) dynamics, using velocity verlet, options: velocity scaling, shake, fixing part of molecule, periodic (cubic) boundary condition. Coordinate from dynamics are compatible with QUANTA (display only!) (v) free energy perturbation and thermodynamic integration. (vi) analysis: rms calculations, internal coordinates LESs standard features: (i) Application of the Locally Enhanced Sampling (LES) method to diffusion, conformational searches and free energy calculations (through specific options in molecular generation). In this approach a number of copies of a small part of the the system (e.g. a number of ligands diffusing through a single protein matrix) are used in order to enhance the sampling of a small part of prime interest. The ligands do not see each other and the protein sees the average force of all the ligands. See Roitberg and Elber, J.Chem.Phys. 95,9277(1991) for current implementation and for modeling side chain positions. (ii) Application of the Self Penalty Walk (SPW) method for reaction path calculations in large molecular systems. In this approach the complete path is represented by a set of discrete points that resemble a chain and the energy of the chain is minimized. See Czerminski and Elber Int.J.Quantum Chem. 24,167(1990) and Nowak, Czerminski and Elber, J.Amer.Chem.Soc. 113,5627(1991) (iii) Calculation of free energy profiles along numerically derived reaction coordinates. Two options: umbrella sampling and free energy perturbation. See Elber, J.Chem.Phys. 93,4312(1990) and Verkhivker and Elber J.Amer.Chem.Soc. in press. This code is not copyrighted and you may transfer it to anyone you like. However the code is provided with the friendly understanding (we simply trust you folks!) that: (a) You will not try to make profit from this code or any parts of it. (b) Any code that you developed based on MOIL will be also available free of charge. (c) In the case that the code is transfered to other users, this message should be included and changes that were made to the code should be stated to the new users. (d) Reference should be made to the authors of the program as listed in moil.src.003/makefile to the extent that this program helped you in your research. Contact Ron Elber (ron@pap.chem.uic.edu) for proper references. MOIL is keyboard based and aimed mostly at developers than at the casual users. MOIL cannot replace any mouse driven commercial program. The program can be copied via anonymous ftp by ------------------------- ftp 128.248.186.70 user-name:anonymous passwd: your email (please) cd dist get moil.003.tar.Z (include only source code and input for test cases total size approx. 2MB) get moil.tests.tar.Z (include input and output files for test cases approx. 6MB) quit ------------------------- On your UNIX site type (we support Stardent-3000, SGI 4D, IBM RS/6000, HP 720) ------------------------ uncompress moil.003.tar.Z tar xvf moil.003.tar repeat the extraction process for the test directory (if you ftp them too) cd export3/moil.src.003 make You may need to uncomment lines in the makefile for your machine. For Stardnet type make_strdnt ------------------------ You will need around 20MB (depending on the machine) to complete the port, the compilation, and the tests. The program will be in the directory export3 the following directories should be below it: moil.src.003 ; moil.mop ; moil.doc ; moil.tests ; tutorial If you do not have access to ftp, or 2MB is too much to transfer send a tape (we can write 1/4" cartridge on Stardent, Sun, LOW DENSITY SGI, 4mm on HP 720 and 1/2" tar tapes) and a self addressed envelope to Carlos Simmerling c/o Prof. Ron Elber Department of Chemistry M/C 111 University of Illinois at Chicago P.O. Box 4348 Chicago IL 60680 From chemistry-request@ccl.net Mon Jul 20 03:34:46 1992 To: chemistry@ccl.net Subject: Neural Nets summary (warrning - LONG STUFF) Date: Mon, 20 Jul 92 01:54:02 -0400 From: radomski@mond1.ccrc.uga.edu Status: R Hello Netters, here goes summary and my own answers to questions about artificial neural networks. I took me a little while but then the Jan's ban put on chemistry@ccl.net was quite a something! BOOKS: "A Practical Quide to Neural Networks" by Marilyn McCord Nelson & W.T. Illingworth both of Texas Instruments, Addison-Wesly 1991 (this is fourt ed, the first one was in 1989) ISBN 0-201-52376-0 contains demo disk for 286-486 PCs with color VGA literature refs. up to about mid 1989 VERY good starting point but not much more "Parallel Distributed Processing" Vols I & II ed. by D.E. Rumelhart & J.L. McClelland of PDP Research Group, UCSD MIT Press 1986 (mine is the first issue, republished uncouth times) ISBN 0-262-18120-7 Vol I ISBN 0-262-13218-4 Vol II ISBN 0-262-18123-1 set of I & II This is the BIBLE of ANN "Explorations in Parallel Distributed Processing" by the same to guys as above, MIT Press 1988 ISBN 0-262-63113-X So called Vol III of R&McC, contains two disks of SOURCE code in C for very many different ANN architectures, and plenty of examples, code runs on PCs and UNIX platforms (oryginal 1988's Ver 1.0 has some bugs, most of which were fixed in Ver 1.1) The MUST for beginner and moderately advanced user alike "Neural Computing, Theory and Practice" by P.D. Wasserman of ANZA Res. Inc. Van Nostrand Reinhold, NY 1989 ISBN 0-442-20743-3 good book esp. for somebody with interest in different architectures (they are all compared from the BP [back propagation] perspective) "Adaptive Pattern Recognition & Neural Networks" by Yoh-Han Pao of Case Western Reserve Univ. Addison-Wesly 1989 ISBN 0-201-12584-6 This is only a selction from the sea of possibilities (quite biased). For examples of ANN usage in chemistry (albeit pretty random ones) good starting point is to browse through special issue of Tetrahedrom Computer Methodology , Vol 3, No 3/4, 1990 which contains also three disks with additional material and demo software (I forget of which firm exactly but it was most probably NeuralWare). BBS', MAILING LISTS & so on: FTP site at cattell.psych.upenn.edu contains back issues od Neuron-Digest, excellent source on all sorts of subjects connected with neural networks (mainly artificial, but not only). The archive contains close to 3000 pages (my estimate) of stuff, heavy going but well worth it!!! Neuron-Digest is the moderated by Peter Marvit. Sam Milosevich has tipped me on some other sources, all on USENET, but since our site is not configured properly to use USENET, I wasn't able to check them: From: sam@ncsa.uiuc.edu (Sam Milosevich) USEnet news conferences in ai: comp.ai comp.ai.edu comp.ai.neural-nets comp.ai.nlang-know-rep comp.ai.philosophy comp.ai.shells comp.ai.vision SOFTWARE: there exists such cornucopia of options that I won't like to waste bandwidth for listing them all. Probably the best way is to browse through back volumes of Neuron-Digest. That is, if one is dissatisfied with, say, software supplied together with the books I mentioned. Regards, Jan --------------- PS. In my oryginal posting I've asked: >>What is the place to look if one is seeking further employment >>in this area? For the last few months there is particular dearth >>of any kind of computational-chemistry advertisements in the >>Science, Nature, Ch&E News and such. Am I overlooking some >>other obvious choice? sadly, there was not a single answer to this. Are we hit by recesion this hard? Jan P. Radomski Complex Carbohydrate Research Center, University of Georgia, 220 Riverbend Rd., Athens, 30602 GA. Ph. (706) 542-4452 Fax (706) 542-4412 e-mail: radomski@mond1.ccrc.uga.edu From chemistry-request@ccl.net Mon Jul 20 06:59:57 1992 Date: Mon, 20 Jul 92 11:05:35 +0200 From: Anders Sundin To: chemistry@ccl.net Subject: CNDO question Status: R I am posting this question for a friend. Can one obtain atomic orbital coefficients for the frontier orbitals from CNDO-calculations? ---- Anders Sundin e-mail: sundinKC@dna.lth.se Organic Chemistry 2 Anders.Sundin@orgk2.lth.se Lund Institute of Technology ok2aps@seldc52.bitnet P.O. Box 124 phone: +46 46 108214 S-22100 Lund, Sweden fax: +46 46 108209 From chemistry-request@ccl.net Mon Jul 20 14:59:51 1992 To: chemistry@ccl.net From: "Larry Wier" Date: 20 Jul 92 11:33:17 DST Subject: FORTRAN 77 compiler Status: R In an attempt to learn more about computational chemistry, I (with help from a friend in Computer Science) recently tried to install MindTool and Psi88 on a Sun SPARC station, but encountered problems when we found that there was not a FORTRAN 77 compiler on the system. Any recommendations on a low cost, FORTRAN 77 compiler that will run on the Suns? Please send replies directly to me and I will summarize for the list. Thank you. Larry Wier ============================== Dr Larry Wier Dept of Chemistry St Bonaventure University St Bonaventure, N.Y. 14778 (716)375-2116 INTERNET:lwier@sbu.edu ============================== From chemistry-request@ccl.net Mon Jul 20 16:55:07 1992 Date: Mon, 20 Jul 92 12:37:02 -0400 From: friedman@tammy.harvard.edu (Dawn Friedman) To: chemistry@ccl.net Subject: QUANTA/MO plots Status: R Can a QUANTA (MS/Polygen) user or guru tell me if there is a QUANTA version which can plot molecular orbital occupation contours? I believe there was some approximation of this feature in Q 2.1. I'd like to be able to send Gaussian MO calculation jobs from QUANTA and retrieve printable plots of any orbital desired. Is this possible? --Dawn From chemistry-request@ccl.net Mon Jul 20 19:54:10 1992 Date: 20 Jul 1992 16:40 -0400 (EDT) From: "Gary D. Anderson" Subject: E-mail address request To: chemistry@ccl.net Status: R Does anyone have an e-mail address for N. N. Greenwood or A. Earnshaw at the University of Leeds, UK? A friend without net access needs to contact them with regard to a question regarding information in "Chemistry of the Elements". Please send replies to m043005@muvms3.mu.wvnet.edu Thanks, Gary D. Anderson Professor of Chemistry Marshall University Huntington, WV From chemistry-request@ccl.net Mon Jul 20 21:53:19 1992 Date: Mon, 20 Jul 1992 16:20:39 -0700 (PDT) From: F0418950@JAGUAR.CSC.WSU.EDU Subject: mopac.man summary To: chemistry@ccl.net Status: R Thank you to all that responded to my request for a postscript version of the mopac6.0 manual. To those that sent me requests for the information I recieved, here's the lineup: I received no word about the mopac manual in postscript form!! This is a bit disheartening, but there are a couple of alternatives. Leonore Findsend and the University of Toledo kindly made available a WriteNow file for the Mac and a reformatted plain text file. Both are in binhex'ed self-extracting archives in the mopac directory on ftp.utoledo.edu. John Simmie of University College Galway said that he had a very spiffy version for LaTEX and that James Stewart might be prodded to make this version available to others( unfortunately I don't have TeX or LaTex, but I'm sure many of you out there do). That's the wrap. - - - Andrew