From chemistry-request@server.ccl.net Tue Sep 28 03:37:40 1999 Received: from sun.fqspl.com.pl. ([212.244.147.3]) by server.ccl.net (8.8.7/8.8.7) with SMTP id DAA00303 for ; Tue, 28 Sep 1999 03:37:35 -0400 Received: from victor by sun.fqspl.com.pl. (SMI-8.6/SMI-SVR4) id JAA19931; Tue, 28 Sep 1999 09:23:56 +0200 Message-ID: <008901bf0983$10065740$1793f4d4@victor.fqspl.com.pl> From: "Victor Anisimov" To: Subject: Re: CCL:mopac7, eps Date: Tue, 28 Sep 1999 09:28:30 +0200 X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Dear Nuran, COSMO does not require any special keywords more than EPS and NSPA to be initiated. But as far as I can judge MOPAC7 can not perform COSMO calculations for excited states. It is implemented only in MOPAC2000. Sincerely, Victor. ========================================================================= Victor Anisimov, PhD, Software Researcher - Computational Chemistry FQS Poland, Palac Pugetow, ul. Starowislna 13-15, 31-038 Krakow, Poland Email: victor@fqspl.com.pl Tel.(+48 12) 429 4345 Fax(+48 12) 429 6124 ========================================================================= -----Original Message----- From: Nuran Elmaci :Dear CCL Members, : :I would like make some excited state calculations including solvent :effects with MOPAC7, I have used EPS=n.nn and NSPA=n keywords, but the :keywords are unrecognized. Is there any other keyword that I should use :with them? There is a description about COSMO in the Mopac manual but :the necessary keywords are not located in the keyword list. I would :appreciate if someone can help me. : :thanks, : :Nuran Elmaci From chemistry-request@server.ccl.net Tue Sep 28 05:36:48 1999 Received: from relay0.unizar.es (tozal.unizar.es [155.210.3.20]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id FAA00952 for ; Tue, 28 Sep 1999 05:36:37 -0400 Received: from [155.210.88.109] (vicort.unizar.es [155.210.88.109]) by relay0.unizar.es (8.9.3/8.9.1) with ESMTP id LAA04290 for ; Tue, 28 Sep 1999 11:30:14 +0200 (MET DST) Message-Id: In-Reply-To: <3.0.5.32.19990927160829.01282c90@wesleyan.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 28 Sep 1999 11:30:06 +0200 To: CHEMISTRY@ccl.net From: Javier Modrego Subject: Re: CCL:Power Mac G4, linux and bechmarking At 16:08 -0400 27/9/99, Yongxing Liu wrote: >Dear CCLers, > >I would like to know which version of linux is optimized for >processors. Does any one have built super clusters with G4? I also want to >know bechmarking on G4/linux for computational chemistry programs (e.g. >Charmm, Amber or g2k/g98 etc). > There is some impresive information about Power Mac G4 clusters with some benchmarks in http://exodus.physics.ucla.edu/appleseed/appleseed.html . In this page there is also links to others PowerMac Clusters. I dont know at the moment about the status of Linux for G4 but you can check at http://www.linuxppc.org , http://www.linuxppc.com/ and http://www.mklinux.apple.com/ Probably an optimized version for G4 has not yet been released. It would be nice to know about the performance of the AltiVec unit. I wonder if somebody has ported Gaussian to PowerMac Linux (G3). Perhaps could share the makefiles... Best wishes, Javier Dr. F.J. Modrego Department of Inorganic Chemistry Facultad de Ciencias University of Zaragoza 50071 ZARAGOZA SPAIN Tel <34>-976-762288 Fax <34>-976-761187 E-mail: modrego@posta.unizar.es From chemistry-request@server.ccl.net Tue Sep 28 07:34:43 1999 Received: from pinar.csic.es (cc.csic.es [161.111.10.10]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id HAA02296 for ; Tue, 28 Sep 1999 07:34:29 -0400 Received: from pinar.csic.es by pinarvms.csic.es (PMDF V5.1-4 #13290) id <01JGI7JKFFOW000BVB@pinarvms.csic.es> for chemistry@www.ccl.net; Tue, 28 Sep 1999 13:31:40 CE X400-Received: by /PRMD=iris/ADMD=mensatex/C=es/; Relayed; Tue, 28 Sep 1999 13:27:09 +0000 Date: Tue, 28 Sep 1999 13:27:09 +0000 From: Javier Ramos Subject: ADF FREQUENCIES. To: chemistry@server.ccl.net Message-id: <256*imtr170@pinar1.csic.es> Content-identifier: 256 MIME-version: 1.0 (Generated by Ean X.400 to MIME gateway) X400-Content-type: P2-1984 (2) X400-MTS-identifier: [/PRMD=iris/ADMD=mensatex/C=es/;990928132709] X400-Originator: imtr170@pinar1.csic.es X400-Recipients: non-disclosure:; Dears CCL's I have recently performed a frequency calculation over a transition structure using the ADF package. I would like to visualizate the vibrational mode which corresponding to the imaginary frequency. My question is: Using the output generated by ADF. How can i visualizate this vibrational mode? Thanks in advance. Javier Ramos imtr170@pinar1.csic.es . From chemistry-request@server.ccl.net Tue Sep 28 10:03:55 1999 Received: from theory.tc.cornell.edu (THEORY.TC.CORNELL.EDU [128.84.30.174]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id KAA02880 for ; Tue, 28 Sep 1999 10:03:55 -0400 Received: from ren.tc.cornell.edu (REN.TC.CORNELL.EDU [128.84.20.129]) by theory.tc.cornell.edu (8.8.4/8.8.3/CTC-1.0) with SMTP id JAA18299 for ; Tue, 28 Sep 1999 09:57:35 -0400 Sender: richard@tc.cornell.edu Message-ID: <37F0C950.2781@tc.cornell.edu> Date: Tue, 28 Sep 1999 09:57:36 -0400 From: Richard Gillilan X-Mailer: Mozilla 3.01SC-SGI (X11; I; IRIX 6.2 IP22) MIME-Version: 1.0 To: chemistry@ccl.net Subject: C++ vs Fortran performance References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > Also, if speed is what is sought-for here--optimized C is just as fast as > fortran. I write all my code in C++ nowadays, but the most frequently asked question I get is "isn't C++ much slower than Fortran?". I have a paper on NAMD by the Schulten group that presents benchmarks comparing their optimized C++ code with similar Fortran codes(*). On a single processor their code takes about 20-30% longer (304 min vs 237-255 min). I have not been able to compare my own code with similar commercial versions yet, but I think this is a very acceptable hit considering all that is gained from C++ and OOP. Has anyone had a similar experience? The paper also gives a few guidelines: o use ordinary C for heavy numerical tasks (simple constructs) o use in-lining for critical functions o avoid virtual functions (runtime function lookups) I have found these to be VERY important. If you have additional guidlines, I would love to collect and summarize for the list. (*) International Journal of Supercomputer Applications and High Performance Computing Vol 10, No 4 Winter 1996 pp251-268 From chemistry-request@server.ccl.net Tue Sep 28 10:37:28 1999 Received: from hermes.bmc.uu.se (hermes.bmc.uu.se [130.238.39.159]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id KAA03077 for ; Tue, 28 Sep 1999 10:37:27 -0400 Received: from alpha2.bmc.uu.se ([130.238.37.65]) by hermes.bmc.uu.se (post.office MTA v2.0 0813 ID# 0-17733) with SMTP id AAA11218; Tue, 28 Sep 1999 15:32:35 +0100 Received: from werkman.bmc.uu.se by alpha2.bmc.uu.se; (5.65v3.2/1.1.8.2/30Jan97-1111AM) id AA15105; Tue, 28 Sep 1999 15:31:05 +0100 Date: Tue, 28 Sep 1999 22:49:12 +0200 (CEST) From: David van der Spoel X-Sender: spoel@werkman.bmc.uu.se To: Richard Gillilan Cc: chemistry@ccl.net Subject: Re: CCL:C++ vs Fortran performance In-Reply-To: <37F0C950.2781@tc.cornell.edu> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 28 Sep 1999, Richard Gillilan wrote: >> Also, if speed is what is sought-for here--optimized C is just as fast as >> fortran. > >I write all my code in C++ nowadays, but the most frequently asked >question I get >is "isn't C++ much slower than Fortran?". I have a paper on NAMD by >the Schulten group that presents benchmarks comparing their optimized >C++ code with similar Fortran codes(*). On a single processor their code >takes about 20-30% longer (304 min vs 237-255 min). This is a recurring topic. The answer depends on application and compiler and can thus not be answered unequivocally. The GROMACS MD code (http://md.chem.rug.nl/~gmx) which was developed by me and others in the Berendsen group, contains code in fortran and C that is identical. Yet the fortran version is always faster, except when using g77 on Linux/x86 (and other platforms probably). This was tested on SGI, DEC alpha, IBM SP2 and on Linux using the Portland compilers. The raw performance of identical code written in different languages is: f77 >= C >= C++. If you are able to concentrate the number crunching in a small fraction of the code, it may be worthwhile to go for mixed languages, like we did with GROMACS. It is 99% C, and 1% fortran, combining the fortran speed and the ease of development of C. Groeten, David. ________________________________________________________________________ Dr. David van der Spoel Biomedical center, Dept. of Biochemistry s-mail: Husargatan 3, Box 576, 75123 Uppsala, Sweden e-mail: spoel@xray.bmc.uu.se www: http://zorn.bmc.uu.se/~spoel phone: 46 18 471 4205 fax: 46 18 511 755 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From chemistry-request@server.ccl.net Tue Sep 28 10:40:11 1999 Received: from mailserv.mta.ca (root@mailserv.mta.ca [138.73.101.5]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id KAA03154 for ; Tue, 28 Sep 1999 10:40:11 -0400 Received: from sage.mta.ca (smtp.mta.ca [138.73.101.97]) by mailserv.mta.ca (8.9.3/8.9.3) with ESMTP id LAA01547 for ; Tue, 28 Sep 1999 11:33:51 -0300 (ADT) Received: from m0612.mta.ca (host-22-237.mta.ca [138.73.22.237]) by sage.mta.ca (8.9.3/8.9.3) with SMTP id LAA25487 for ; Tue, 28 Sep 1999 11:37:20 -0300 (ADT) Message-Id: <3.0.5.32.19990928113843.007ce100@mailserv.mta.ca> X-Sender: mculf@mailserv.mta.ca X-Mailer: QUALCOMM Windows Eudora Light Version 3.0.5 (32) Date: Tue, 28 Sep 1999 11:38:43 -0300 To: chemistry@ccl.net From: Miroslava Culf Subject: Lipid bilayer simulations - summary Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" My original question was: > I would like to find out more about computer simulations of lipid bilayer. > Thus I would appreciate any information about RECENT review articles or > books dealing with any aspect of computer simulation of lipid bilayer or > behaviour of lipids in the bilayer. Thanks very much to David van der Spoel, Ramesh V. Durvasula and Rick Venable. There suggestions were: David van der Spoel: Check out the articles from the Berendsen group: http://md.chem.rug.nl/articles.html Especially for authors Tieleman and Marrink Ramesh V. Durvasula: You should contact the lab of Prof. Ching-hsien Huang, Dept of Biochemistry, Univ. of Virginia. He is one of the foremost authority on lipid bilayers, and his lab has done quite a bit of work on bilayer simulations. His email address is ch9t@virginia.edu. Rick Venable: Our group has worked extensively in this area, and we have a chapter in the book "Biological Membranes: A Molecular Perspective from Computation and Experiment", edited by Kenneth M Merz, Jr., and Benoit Roux, Birkhauser, 1996 ISBN 0-8176-3827-X We also have a number of journal articles; some recent works are Feller SE, Pastor RW. Constant surface tension simulations of lipid bilayers: The sensitivity of surface areas and compressibilities J CHEM PHYS 111: (3) 1281-1287 JUL 15 1999 Feller SE, Venable RM, Pastor RW Computer simulation of a DPPC phospholipid bilayer: Structural changes as a function of molecular surface area LANGMUIR 13: (24) 6555-6561 NOV 26 1997 Feller SE, Yin DX, Pastor RW, et al. Molecular dynamics simulation of unsaturated lipid bilayers at low hydration: Parameterization and comparison with diffraction studies BIOPHYS J 73: (5) 2269-2279 NOV 1997 Sincerely, Mira ____________________________________________________________ Miroslava Culf, Ph.D. 65 York St. Computer Services Department/Chemistry Department Mount Allison University Sackville, NB E4L 1E4 Canada e-mail: mculf@mta.ca tel. 506-364-2411 http://www.mta.ca/faculty/science/chem/staff/mculf.htm ____________________________________________________________ From chemistry-request@server.ccl.net Tue Sep 28 11:24:37 1999 Received: from ns.apbiotech.com (ns.apbiotech.com [192.36.117.130]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id LAA03398 for ; Tue, 28 Sep 1999 11:24:36 -0400 From: enrique.carredano@eu.apbiotech.com Received: from eunotes.apbiotech.com (unverified [193.180.20.124]) by ns.apbiotech.com (Rockliffe SMTPRA 3.3.1) with SMTP id for ; Tue, 28 Sep 1999 17:19:35 +0200 Received: by eunotes.apbiotech.com(Lotus SMTP MTA v4.6.3 (778.2 1-4-1999)) id 412567FA.00592301 ; Tue, 28 Sep 1999 17:13:36 +0100 X-Lotus-FromDomain: XYNET To: Chemistry@ccl.net Message-ID: <412567FA.005921A5.00@eunotes.apbiotech.com> Date: Tue, 28 Sep 1999 17:15:12 +0100 Subject: CCL:protein data Mime-Version: 1.0 Content-type: text/plain; charset=us-ascii Content-Disposition: inline Dear CCLrs, Does anyone know where to find tables of data relevant to different proteins, namely diffusion coefficients, extinction coefficients and isoelectric points? More specifically, my questions are: What is the diffusion coefficient (in aquoeus solutions) for these proteins? What is their extinction coefficients (at 280 nm)? What is the isoelectric point for Catalase? |-----------------+------------------------| | | | | Proteins | Source | | | | |-----------------+------------------------| | | | | a-lactalbumin | bovine milk | | | | |-----------------+------------------------| | | | | myoglobulin | horse skeletal | | | | |-----------------+------------------------| | | | | trypsinogen | bovine pancreas | | | | |-----------------+------------------------| | | | | pepsin | porcine stomach mucosa | | | | |-----------------+------------------------| | | | | ovalbumin | chicken egg | | | | |-----------------+------------------------| | | | | BSA | bovine | | | | |-----------------+------------------------| | | | | lactoferrin | bovine | | | | |-----------------+------------------------| | | | | IgG | human | | (polyclonal) | | | | | |-----------------+------------------------| | | | | catalase | bovine liver | | | | |-----------------+------------------------| | | | | thyroglobulin | porcine | | | | |-----------------+------------------------| As usual, I will sumarize the answers if any. Thanks Enrique Enrique Carredano, MSc, PhD +46 (0)18 16 50 00 tel Polymer and Surface Chemistry +46 (0)18 16 50 42 direct Amersham Pharmacia Biotech +46 (0)18 16 63 96 fax Bjorkg 30, 751 84 Uppsala Sweden enrique.carredano@eu.apbiotech.com From chemistry-request@server.ccl.net Tue Sep 28 11:36:47 1999 Received: from mserve1.acs.ucalgary.ca (mserve1.acs.ucalgary.ca [136.159.34.51]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id LAA03463 for ; Tue, 28 Sep 1999 11:36:47 -0400 received: from acs1.acs.ucalgary.ca(136.159.34.221) by mserve1.acs.ucalgary.ca via smap (V2.0) id ZZ474921; Tue, 28 Sep 99 09:27:56 -0600 Received: (from patchkov@localhost) by ucalgary.ca (AIX4.3/UCB 8.8.8/8.8.8) id JAA27298; Tue, 28 Sep 1999 09:27:50 -0600 Message-Id: <199909281527.JAA27298@ucalgary.ca> Subject: Re: CCL:ADF FREQUENCIES. To: imtr170@pinar1.csic.es (Javier Ramos) Date: Tue, 28 Sep 99 9:27:49 MDT From: "Serguei Patchkovskii" Cc: chemistry@server.ccl.net In-Reply-To: ; from "Javier Ramos" at Sep 28, 1999 1:27 pm X-Mailer: ELM [version 2.3 PL11K] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > Using the output generated by ADF. How can i visualizate this > vibrational mode? There are many ways of doing this. The one I use is to generate an XMOL (http://www.msc.edu/msc/docs/xmol/XMol.html) movie file from the ADF vibrational analysis output using the awk script attached at the end of this message. For example, if your output file is called 'moi-molecule.log', and you want to visualize the mode with the vibrational frequency of 1234.01 cm^-1, you'll say: awk -f adf2vib.awk moi-molecule.log freq=1234.01 >mode1234.xyz xmol -readFormat xyz mode1234.xyz ... and use xmol control panels to start the animation. The script works fine for ADF 2.3.*; it should be trivial to modify it for any other version of ADF. Regards, /Serge.P --- cut here: adf2vib.awk BEGIN { infinity = 10000000 ; COORDRecord = infinity ; MODERecord = infinity ; natoms = 0 ; iatoms = 0 ; freqOn = 0 ; modeOn = 0 ; snapshots = 20 ; scale = 3.0 ; } NR==1 && (freq=="") { print "Missing freq= on the command line" ; exit 1 ; } /\* F R E Q U E N C I E S \*/ { freqOn = 1 ; } /^ *Coordinates \(Cartesian\)$/ && (freqOn==1){ COORDRecord = NR ; natoms = 0 ; next ; } NR == COORDRecord + 1 && !/^ *==*$/{ COORDRecord = infinity ; next ; } NR == COORDRecord + 3 && !/ *Atom *bohr *angstrom/{ COORDRecord = infinity ; next ; } NR == COORDRecord + 4 && !/ *X *Y *Z *X *Y *Z/{ COORDRecord = infinity ; next ; } NR == COORDRecord + 5 && !/^ --*$/{ COORDRecord = infinity ; next ; } NR >= COORDRecord + 6 { if( NF < 8 ){ COORDRecord = infinity ; next ; } label[natoms] = $2 ; x [natoms] = $6 ; y [natoms] = $7 ; z [natoms] = $8 ; natoms++ ; } /Vibrations and Normal Modes.*cartesian coordinates, NOT mass-weighted/{ modeOn = 1 } /Normal Modes in Z-matrix coordinates/{ modeOn = 0 } (NF<=3) && (modeOn==1) { for( i = 1 ; i <= NF ; i++ ){ if( $i == freq ){ ndisp = 0 ; MODERecord = NR ; field = i ; next ; } } } (NR >= MODERecord + 2) && NF < 1 { MODERecord = infinity ; reportVib() ; exit 0 } (NR >= MODERecord + 2) { alabel = substr($1,index($1,".")+1) ; if( alabel != label[ndisp] ){ printf "Atom %d has different labels in XYZ (%s) and displacement (%s)\n", \ ndisp+1, label[ndisp], alabel ; exit 1 ; } dispx[ndisp] = $(2+3*(field-1)) ; dispy[ndisp] = $(3+3*(field-1)) ; dispz[ndisp] = $(4+3*(field-1)) ; ndisp++ ; } function dumpgeo() { print natoms ; print "" ; for( i = 0 ; i < natoms ; i++ ){ printf "%4s %20.8f %20.8f %20.8f\n", label[i], x[i], y[i], z[i] ; } } function reportVib() { if( ndisp != natoms ){ printf "Number of atoms (%d) and displacement vectors (%d) do not agree\n", \ natoms, ndisp ; exit 1 ; } for( iatom = 0 ; iatom < natoms ; iatom++ ){ rx[iatom] = x[iatom] ; ry[iatom] = y[iatom] ; rz[iatom] = z[iatom] ; } for( isnap = -snapshots ; isnap <= snapshots ; isnap ++ ){ s = (scale*isnap)/snapshots ; for( iatom = 0 ; iatom < natoms ; iatom++ ){ x[iatom] = rx[iatom] + s*dispx[iatom] ; y[iatom] = ry[iatom] + s*dispy[iatom] ; z[iatom] = rz[iatom] + s*dispz[iatom] ; } dumpgeo() ; } } --- cut here: end of adf2vib.awk -- home page: http://www.cobalt.chem.ucalgary.ca/ps/ From chemistry-request@server.ccl.net Tue Sep 28 12:22:21 1999 Received: from ice.lanl.gov (ice.lanl.gov [128.165.22.6]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id MAA03698 for ; Tue, 28 Sep 1999 12:22:20 -0400 Received: from localhost (localhost [[UNIX: localhost]]) by ice.lanl.gov (8.9.3/8.9.3) id KAA18356; Tue, 28 Sep 1999 10:14:48 -0600 From: Matt Challacombe Reply-To: MChalla@T12.LANL.Gov Organization: Group T-12, Theoretical Chemistry and Molecular Physics To: David van der Spoel , Richard Gillilan Subject: Re: CCL:C++ vs Fortran performance Date: Tue, 28 Sep 1999 10:08:46 -0600 X-Mailer: KMail [version 1.0.24] Content-Type: text/plain Cc: chemistry@ccl.net References: MIME-Version: 1.0 Message-Id: <9909281014480K.12903@ice.lanl.gov> Content-Transfer-Encoding: 8bit Optimization in C (and C++) is NP hard. Optimization in Fortran is polynomial. Thus, the difference between Fortran and C and their various dialects becomes more apparent as code complexity and compile time increases. -Matt +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + Matt Challacombe, Ph.D. http://www.t12.lanl.gov/~mchalla/ + + Los Alamos National Laboratory email: mchalla@t12.lanl.gov + + Theoretical Division vmail: (505) 698-4112 + + Group T-12, Mail Stop B268 phone: (505) 665-5905 + + Los Alamos, New Mexico 87545 fax: (505) 665-3909 + + + + "The secret to mountain biking is pretty simple. The slower you go + + the more likely it is you'll crash." -- Julie Furtado + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ From chemistry-request@server.ccl.net Tue Sep 28 12:31:35 1999 Received: from comsig.nibsc.ac.uk (comsig.nibsc.ac.uk [193.62.43.13]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id MAA03798 for ; Tue, 28 Sep 1999 12:31:33 -0400 Received: from nibsc.ac.uk (dlinmf.nibsc.ac.uk [193.62.42.144]) by comsig.nibsc.ac.uk (980427.SGI.8.8.8/970903.SGI.AUTOCF) via ESMTP id RAA13787 for ; Tue, 28 Sep 1999 17:24:43 +0100 (BST) Message-ID: <37F0EA86.432C56F5@nibsc.ac.uk> Date: Tue, 28 Sep 1999 17:19:18 +0100 From: Mark Forster Organization: NIBSC X-Mailer: Mozilla 4.05 [en] (Win95; I) MIME-Version: 1.0 To: chemistry@server.ccl.net Subject: Tinker Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear CCL readers Can someone offer some description of how the Tinker suite of codes, in particular the executable pdbxyz.x, assign a force field atom type for an atom in a particular pdb file ? In general I see there to be two ways to do this for any force field code. (i) A general scheme based upon element types in particular bonding situations - this would be harder to code easier to apply. (ii) An easier to code scheme in which an atom name and residue name map to a force field atom type. As an extension of this how could one add ones own parameter set so as to study new functional groups with Tinker ? I will summarize any replies if they are not already posted to CCL. Mark -- Dr Mark J Forster Ph.D. Principal Scientist Informatics Laboratory National Institute for Biological Standards and Control Blanche Lane, South Mimms, Hertfordshire EN6 3QG, United Kingdom. Tel +44 (0)1707 654753 FAX +44 (0)1707 646730 E-mail mforster@nibsc.ac.uk From chemistry-request@server.ccl.net Tue Sep 28 05:36:21 1999 Received: from bacchus.pc1.uni-duesseldorf.de (root@bacchus.pc1.uni-duesseldorf.de [134.99.152.11]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id FAA00940; Tue, 28 Sep 1999 05:36:19 -0400 Received: from bacchus.pc1.uni-duesseldorf.de (jochen@localhost [127.0.0.1]) by bacchus.pc1.uni-duesseldorf.de (8.8.7/8.8.7) with SMTP id LAA26861; Tue, 28 Sep 1999 11:29:09 +0200 From: Jochen Reply-To: jochen@pc1.uni-duesseldorf.de Organization: Heinrich-Heine-Universität To: Richard Gillilan , CCL Subject: Re: CCL:C to Fortran Date: Fri, 4 Nov 2135 17:56:04 +0200 X-Mailer: KMail [version 1.0.28] Content-Type: text/plain Cc: CHEMISTRY@ccl.net References: <37EF7312.2781@tc.cornell.edu> In-Reply-To: <37EF7312.2781@tc.cornell.edu> MIME-Version: 1.0 Message-Id: <99092811290900.26649@bacchus.pc1.uni-duesseldorf.de> Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id FAA00941 On Mon, 27 Sep 1999 Richard Gillilan wrote: >It is possible to call Fortran routines from C programs (I have done it >in the past). It is also possible to call C from Fortran. On SGI >platforms, >there is a program called "mkf2c" to create the appropriate wrappers >for C code. There is the cfortran package to do that portably. -- Jochen Heinrich-Heine-Universität, Institut für Physikalische Chemie I Universitätsstr. 1, Geb. 26.43.02.29, 40225 Düsseldorf, Germany phone 02118113681 fax 02118115195 -- www-public.rz.uni-duesseldorf.de/~jochen Jochen@Uni-Duesseldorf.de -- Jochen.Kuepper@FernUni-Hagen.de -- Kuepper@ACM.org From chemistry-request@server.ccl.net Tue Sep 28 05:39:13 1999 Received: from bacchus.pc1.uni-duesseldorf.de (root@bacchus.pc1.uni-duesseldorf.de [134.99.152.11]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id FAA00980 for ; Tue, 28 Sep 1999 05:39:11 -0400 Received: from bacchus.pc1.uni-duesseldorf.de (jochen@localhost [127.0.0.1]) by bacchus.pc1.uni-duesseldorf.de (8.8.7/8.8.7) with SMTP id LAA26868; Tue, 28 Sep 1999 11:31:32 +0200 From: Jochen Reply-To: jochen@pc1.uni-duesseldorf.de Organization: Heinrich-Heine-Universität To: Matt Challacombe , Iraj Daizadeh , CHEMISTRY@ccl.net Subject: Re: CCL:C to Fortran Date: Tue, 28 Sep 1999 11:30:50 +0200 X-Mailer: KMail [version 1.0.28] Content-Type: text/plain References: <9909271410470G.12903@ice.lanl.gov> In-Reply-To: <9909271410470G.12903@ice.lanl.gov> MIME-Version: 1.0 Message-Id: <99092811313201.26649@bacchus.pc1.uni-duesseldorf.de> Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id FAA00982 On Mon, 27 Sep 1999 Matt Challacombe wrote: >On Mon, 27 Sep 1999, Iraj Daizadeh wrote: > >> Why in the world would someone want to perform this absurd transformation? > >One reason for performing this transformation is to ensure portability >for fortran-c and c-fortran interfaces where strings are passed. As said before, cfortran does the magic :-) IIRC, it even works for strings. -- Jochen Heinrich-Heine-Universität, Institut für Physikalische Chemie I Universitätsstr. 1, Geb. 26.43.02.29, 40225 Düsseldorf, Germany phone 02118113681 fax 02118115195 -- www-public.rz.uni-duesseldorf.de/~jochen Jochen@Uni-Duesseldorf.de -- Jochen.Kuepper@FernUni-Hagen.de -- Kuepper@ACM.org From chemistry-request@server.ccl.net Tue Sep 28 05:43:37 1999 Received: from bacchus.pc1.uni-duesseldorf.de (root@bacchus.pc1.uni-duesseldorf.de [134.99.152.11]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id FAA01015 for ; Tue, 28 Sep 1999 05:43:36 -0400 Received: from bacchus.pc1.uni-duesseldorf.de (jochen@localhost [127.0.0.1]) by bacchus.pc1.uni-duesseldorf.de (8.8.7/8.8.7) with SMTP id LAA26885; Tue, 28 Sep 1999 11:36:58 +0200 From: Jochen Reply-To: jochen@pc1.uni-duesseldorf.de Organization: Heinrich-Heine-Universität To: "Dr.Teerakiat Kerdcharoen" , chemistry@server.ccl.net Subject: Re: CCL:Linux distributions and FORTRAN (g77) Date: Tue, 28 Sep 1999 11:34:52 +0200 X-Mailer: KMail [version 1.0.28] Content-Type: text/plain References: <19990927133440.2952.qmail@nwcst314.netaddress.usa.net> In-Reply-To: <19990927133440.2952.qmail@nwcst314.netaddress.usa.net> MIME-Version: 1.0 Message-Id: <99092811365803.26649@bacchus.pc1.uni-duesseldorf.de> Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id FAA01016 On Die, 28 Sep 1999 Dr.Teerakiat Kerdcharoen wrote: >According to Mike Falcetta's problem on G77 compiler on Linux, >I would like to share some experiences. G77 has no option to >promote single precision to double precision, namely -r8, which 'Cause it's of no need. Write your code the way you want. >is available in most F77 compiler. I also had disastrous results >>from MD simulations without double precision. All we can do is Of, course, what did you expect ? >that (1) use f2c instead (2) manually change all single precision >variables to double that is laborious or (3) obtain free full sed s/real/real*8/ -- Jochen Heinrich-Heine-Universität, Institut für Physikalische Chemie I Universitätsstr. 1, Geb. 26.43.02.29, 40225 Düsseldorf, Germany phone 02118113681 fax 02118115195 -- www-public.rz.uni-duesseldorf.de/~jochen Jochen@Uni-Duesseldorf.de -- Jochen.Kuepper@FernUni-Hagen.de -- Kuepper@ACM.org From chemistry-request@server.ccl.net Tue Sep 28 08:39:17 1999 Received: from ccl.net (atlantis.ccl.net [192.148.249.4]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id IAA02558 for ; Tue, 28 Sep 1999 08:39:17 -0400 Received: from csb0.IPC.PKU.EDU.CN (csb0.ipc.pku.edu.cn [162.105.177.12]) by ccl.net (8.8.6/8.8.6/OSC 1.1) with SMTP id IAA17513 for ; Tue, 28 Sep 1999 08:32:51 -0400 (EDT) Received: by csb0.IPC.PKU.EDU.CN (920330.SGI/940406.SGI.AUTO) for chemistry@www.ccl.net id AA08954; Tue, 28 Sep 99 20:32:22 -0700 Date: Tue, 28 Sep 1999 20:32:21 -0700 (PDT) From: Tao Peng To: ccl Subject: How to call a C++ program from C? Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear All, I have a big program which is written in C and Fortran, but mainly in C. The author compile each module by their own compiler ( cc and f77 on SGI) and link them by f77. Now I have another C++ program (still others) and want incorporate it into the previous big program. My idea is want to call C++ from C, but compile them separately, and link them altogether. Can anyone tell me whether this idea is possible, or have better suggestion? Thank very in advance. Have a nice day!!! Yours Sincerely Tao Peng \*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\* \* \* \* Tao Peng \* \* \* \* Molecular Design Laboratory \* \* Institute of Physical Chemistry \* \* Peking University \* \* Beijing, 100871 \* \* P.R.China \* \* \* \* E-mail:taop@csb0.ipc.pku.edu.cn \* \* Lab Phone: 86-10-62756833 \* \* Fax Number:86-10-62751725 \* \* http://www.ipc.pku.edu.cn/~taop/ \* \* Address:Peking University Building 46 Room 2122 \* \* Dormitory Phone: 86-10-62761206 \* \*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\*\* From chemistry-request@server.ccl.net Tue Sep 28 10:12:25 1999 Received: from ccl.net (atlantis.ccl.net [192.148.249.4]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id KAA02938 for ; Tue, 28 Sep 1999 10:12:25 -0400 Received: from mcmail.cis.McMaster.CA (mcmail.CIS.McMaster.CA [130.113.64.66]) by ccl.net (8.8.6/8.8.6/OSC 1.1) with ESMTP id KAA21405 for ; Tue, 28 Sep 1999 10:06:01 -0400 (EDT) Received: from localhost (mattacf@localhost) by mcmail.cis.McMaster.CA with SMTP id KAA26126 for ; Tue, 28 Sep 1999 10:05:27 -0400 (EDT) Date: Tue, 28 Sep 1999 10:05:24 -0400 (EDT) From: "C.F. Matta" To: CHEMISTRY@www.ccl.net Subject: SUMMARY: Clar's Sextet Theory Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi evreyone, Three weeks ago, I posted the following question: Could anyone point me to the key references about Clar's sextet rule of poly-condensed aromatic hydrocarbons? I appreciate very much the answers I got from Robert Ponec, Visnja Stepanic, and Shridhar Gadre. Thank you all very much. Here are their individual replies: Robert Ponec, Rponec@icpf.cas.cz ================================= I can recommend you looking at the paper of O. Polansky and G. Derflinger in the first issue of Int. J. Quant. Chem. in 1967. The approximate title of the paper is "Zur Clarschen theorie der lokaler benzoider gebeite in kondensierter aromatische kohlewasserstoffen", I do not remember already the precise page but it was in the first issue od IJQCH in 1967) The paper is in German but I believe you can find the reference to original Clar`s works there. Good luck R. Ponec ------------------------------------------------------------------------------ Visnja Stepanic, simek@rudjer.irb.hr ==================================== For example, E. Clar, The Aromatic Sextet, John Wiley and Sons Ltd, 1972. Best regards, Visnja ------------------------------------------------------------------------------- Shridhar Gadre, gadre@unipune.ernet.in ====================================== Dear Cherif : Pl. see a paper by C. H. Suresh and S. R. Gadre in J. Org. Chem. 64, 2505 (1999). Pl. send me a summary of the other replies to your query. Thanks.........Shridhar Gadre -------------------------------------------------------------------------------- Cherif ....................................................................... Cherif F. Matta tel. (905) 525-9140 ext. 22502 Chemistry Department fax (905) 522-2509 McMaster University Hamilton, Ontario, CANADA L8S 4M1. ....................................................................... From chemistry-request@server.ccl.net Tue Sep 28 11:03:10 1999 Received: from kasei.materials.ox.ac.uk (kasei.materials.ox.ac.uk [163.1.64.46]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id LAA03309 for ; Tue, 28 Sep 1999 11:03:08 -0400 Received: from localhost (hfield@localhost) by kasei.materials.ox.ac.uk (8.9.3/8.9.3) with ESMTP id PAA21835 for ; Tue, 28 Sep 1999 15:57:08 +0100 X-Authentication-Warning: kasei.materials.ox.ac.uk: hfield owned process doing -bs Date: Tue, 28 Sep 1999 15:57:08 +0100 (BST) From: Andrew Horsfield X-Sender: hfield@kasei.materials.ox.ac.uk Reply-To: Andrew Horsfield To: Computational Chemistry List Subject: Re: CCL:C++ vs Fortran performance In-Reply-To: <37F0C950.2781@tc.cornell.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 28 Sep 1999, Richard Gillilan wrote: > > Also, if speed is what is sought-for here--optimized C is just as fast as > > fortran. > > I write all my code in C++ nowadays, but the most frequently asked > question I get is "isn't C++ much slower than Fortran?". I have a > paper on NAMD by the Schulten group that presents benchmarks comparing > their optimized C++ code with similar Fortran codes(*). On a single > processor their code takes about 20-30% longer (304 min vs 237-255 > min). I have not been able to compare my own code with similar > commercial versions yet, but I think this is a very acceptable hit > considering all that is gained from C++ and OOP. > > Has anyone had a similar experience? I write all my code in C, and am not aware of any performance bottlenecks introduced by the language. (There is one possible problem associated the fact that a C compiler does not know if two parameters passed to a subroutine share the same piece of memory or not - but even this has been overcome by the new standard). Much more important, I suspect, is the quality of the compiler, and the design of the code. I have found gcc delivers good performance on x86 platforms. Problem that can severely reduce performance (such as cache misses) are common to all languages. If anyone has any experience with Ada, I would be interested to know how this performs. Best regards, Andrew +----------------------------------------------------+ Andrew Horsfield e-mail: horsfield@fecit.co.uk FECIT, 2 Longwalk Road, Stockley Park, Uxbridge, Middlesex UB11 1AB, United Kingdom. phone: +44-(0)181-606-4653 FAX: +44-(0)181-606-4422 +----------------------------------------------------+ From chemistry-request@server.ccl.net Tue Sep 28 13:24:09 1999 Received: from mail.mpi-sb.mpg.de (mail.mpi-sb.mpg.de [139.19.1.1]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id NAA04182 for ; Tue, 28 Sep 1999 13:24:07 -0400 Received: from mpii01707.ag1.mpi-sb.mpg.de (mpii01707.ag1.mpi-sb.mpg.de [139.19.10.20]) by mail.mpi-sb.mpg.de (8.8.7/8.8.8) with ESMTP id TAA05701 for ; Tue, 28 Sep 1999 19:17:43 +0200 Received: from mpii01707 (localhost.mpi-sb.mpg.de [127.0.0.1]) by mpii01707.ag1.mpi-sb.mpg.de (8.8.8/8.8.8) with ESMTP id TAA27638 for ; Tue, 28 Sep 1999 19:17:35 +0200 (MET DST) Resent-Message-Id: <199909281717.TAA27638@mpii01707.ag1.mpi-sb.mpg.de> Message-Id: <199909281717.TAA27638@mpii01707.ag1.mpi-sb.mpg.de> X-Mailer: exmh version 2.0.3 2/24/98 To: Richard Gillilan Subject: Re: CCL:C++ vs Fortran performance In-reply-to: Your message of "Tue, 28 Sep 1999 09:57:36 EDT." <37F0C950.2781@tc.cornell.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 28 Sep 1999 17:47:01 +0200 From: Oliver Kohlbacher Resent-To: chemistry@ccl.net Resent-Date: Tue, 28 Sep 1999 19:17:35 +0200 Resent-From: Oliver Kohlbacher > C++ code with similar Fortran codes(*). On a single processor their code > takes about 20-30% longer (304 min vs 237-255 min). > I have not been able to compare my own code with similar commercial > versions yet, but I think this is a very acceptable hit considering > all that is gained from C++ and OOP. According to my experiences 30% seems to be an average value if you take a bit of care. Thoroughly optimized code can even be faster. There is a paper by Veldhuizen and Jernigan ("Will C++ be faster than fortran", ISCOPE'97, extended abstract) that states that C++ might even be faster. I can only recommend this paper and the stuff related to their Blitz++ library. > The paper also gives a few guidelines: > > o use ordinary C for heavy numerical tasks (simple constructs) > o use in-lining for critical functions > o avoid virtual functions (runtime function lookups) I would add the following rules: o use templates and template specialization. Although template code takes a long time to compile, it is very fast o virtual function can often be replaced using templates o use compact data structures to increase the locality of your code. Data encapsulation can also increases locality. o use the better algorithms and data structures - C++ is much better suited to implement more sophisticated algorithms and data structures than FORTRAN and many algorithms are already available in the libraries Regards, Oliver ---- Oliver Kohlbacher (oliver@mpi-sb.mpg.de) Max-Planck-Institut fuer Informatik, Im Stadtwald, 66123 Saarbruecken/GERMANY Tel.: 0681-9325-505 Fax: 0681-9325-199 From chemistry-request@server.ccl.net Tue Sep 28 16:26:30 1999 Received: from nucleus.harvard.edu (nucleus.harvard.edu [140.247.90.196]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id QAA05499 for ; Tue, 28 Sep 1999 16:26:30 -0400 Received: from localhost (daizadeh@localhost) by nucleus.harvard.edu (8.9.3/8.9.3) with ESMTP id QAA02574; Tue, 28 Sep 1999 16:21:39 -0400 (EDT) Date: Tue, 28 Sep 1999 16:21:39 -0400 (EDT) From: Iraj Daizadeh To: Richard Gillilan , CCL Subject: Re: CCL:C to Fortran In-Reply-To: <99092811290900.26649@bacchus.pc1.uni-duesseldorf.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII This process is trivial and is essentially unix platform independent. See my review in the archives-- or for a most trivial example ... see http://www.mcb.harvard.edu/gilbert/daizadeh/multi_prog.html Iraj. > On Mon, 27 Sep 1999 Richard Gillilan wrote: > > >It is possible to call Fortran routines from C programs (I have done it > >in the past). It is also possible to call C from Fortran. On SGI > >platforms, > >there is a program called "mkf2c" to create the appropriate wrappers > >for C code. > > There is the cfortran package to do that portably. > > -- Jochen Iraj Daizadeh, Ph.D. Harvard University Department of Cellular and Molecular Biology The Biological Laboratories 16 Divinity Avenue Cambridge, MA 02138 Phone: (617) 495-0783 (617) 495-0560 Fax: (617) 496-4313 Email: daizadeh@nucleus.harvard.edu WebPage: http://mcb.harvard.edu/gilbert/daizadeh From chemistry-request@server.ccl.net Tue Sep 28 16:53:55 1999 Received: from swan2.uspnet.usp.br (swan2.uspnet.usp.br [143.107.253.69]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id QAA05730 for ; Tue, 28 Sep 1999 16:52:47 -0400 Received: from p300 (ligquim.ffclrp.usp.br [143.107.194.89]) by swan2.uspnet.usp.br (8.9.3/8.9.3) with SMTP id RAA17962 for ; Tue, 28 Sep 1999 17:39:51 -0300 (EST) Message-ID: <020201bf09f1$b5479980$0100005a@ligquim.ffclrp.usp.br> From: "Sergio Emanuel Galembeck" To: "CCL mail list" Subject: Compiling GAMESS in Linux Date: Tue, 28 Sep 1999 17:40:31 -0300 Organization: Laboratorio de Modelagem Molecular - FFCLRP-USP MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Dear All, Does anyone know how to compile GAMESS in RedHat 5.0, with g77 1.1.2? Best, Sergio From chemistry-request@server.ccl.net Tue Sep 28 18:57:43 1999 Received: from life.ai.mit.edu (life.ai.mit.edu [128.52.32.80]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id SAA07527 for ; Tue, 28 Sep 1999 18:57:43 -0400 Received: from genie.ai.mit.edu (IDENT:grg@genie.ai.mit.edu [128.52.38.40]) by life.ai.mit.edu (8.9.1/8.9.3/AI2.13/ai.master.life:2.10) with ESMTP id SAA03180; Tue, 28 Sep 1999 18:51:07 -0400 (EDT) Received: (from grg@localhost) by genie.ai.mit.edu (8.9.1a/AI2.7/ai.client:2.1) id SAA05388; Tue, 28 Sep 1999 18:51:05 -0400 Message-Id: <199909282251.SAA05388@genie.ai.mit.edu> Subject: Re: CCL:C++ vs Fortran performance To: oliver@mpi-sb.mpg.de (Oliver Kohlbacher), chemistry@ccl.net Date: Tue, 28 Sep 1999 18:51:05 -0400 (EDT) Cc: richard@tc.cornell.edu (Richard Gillilan) In-Reply-To: <199909281717.TAA27638@mpii01707.ag1.mpi-sb.mpg.de> from "Oliver Kohlbacher" at Sep 28, 1999 05:47:01 PM From: grg22@ai.mit.edu X-Mailer: ELM [version 2.5 PL0pre8] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Oliver Kohlbacher wrote: > > o use ordinary C for heavy numerical tasks (simple constructs) > > o use in-lining for critical functions > > o avoid virtual functions (runtime function lookups) > I would add the following rules: > o use templates and template specialization. Although template code > takes a long time to compile, it is very fast > o virtual function can often be replaced using templates > o use compact data structures to increase the locality of your code. > Data encapsulation can also increases locality. > o use the better algorithms and data structures > - C++ is much better suited to implement > more sophisticated algorithms and data structures than FORTRAN and many > algorithms are already available in the libraries Also pretty important for C++ optimization: o use references (the "&" operator) in parameter lists o make liberal use of the "const" keyword, as in: const memberFunction(const Classname &class, int i) const { ... } --grg