From ccl@www.ccl.net Tue Oct 21 04:47:49 1997 Received: from bedrock.ccl.net for ccl@www.ccl.net by www.ccl.net (8.8.3/950822.1) id EAA21828; Tue, 21 Oct 1997 04:06:22 -0400 (EDT) Received: from danpost.uni-c.dk for peon@medchem.dfh.dk by bedrock.ccl.net (8.8.6/950822.1) id EAA01067; Tue, 21 Oct 1997 04:06:21 -0400 (EDT) Received: from medchem.dfh.dk (medchem.dfh.dk [130.225.177.15]) by danpost.uni-c.dk (8.8.6/8.6) with SMTP id KAA25011; Tue, 21 Oct 1997 10:06:19 +0200 (METDST) Received: from [130.225.177.59] (compmac2 [130.225.177.59]) by medchem.dfh.dk (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id KAA18151; Tue, 21 Oct 1997 10:28:20 +0200 Message-Id: In-Reply-To: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 21 Oct 1997 08:58:21 +0200 To: From: Per-Ola Norrby Subject: Re: CCL:PCM calculations.. Cc: chemistry@ccl.net Dear Qiang Cui, >Hi, folks. Got a few questions concerning PCM for solvation effect, >I'd appreciate any comments from the experts ouut there. >2. In a recent CPL paper by Tomasi et al., a new implementation of PCM, >the so-called PCM/DIR has been reported. ANy one has any experience with >it? or at least how to get it? I've been using it, I can recommend it highly. The problem is getting it... The only way I know is to contact Tomasi directly. For a while they gave it away freely, now I'm not so sure. In my experience, it's a lot more accurate than all the PCM implementations in Gaussian94. I have some cases with relative aqueous pKa for ionic systems (not yet published) where PCM/DIR was within a pKa unit whereas SCIPCM was 8 pKa units off. Of the methods we've been interested in, only the Cramer-Truhlar SM5 yielded an accuracy comparable to PCM/DIR. >3. Is there any new development in the gradient calculations with PCM >recently? Or, any other methods has analytical gradients? (besides >Truhlar's g-Born and Troung's GCOSMO) The latest version of PCM/DIR includes analytical gradients. We've tested it, it's OK but not perfectly stable. We're trying it for zwitterionic systems, that may be stretching it a bit... >4. In several papers by Tomasi et al.,the SCIPCM of Frisch et al. was >shown to be not as accurate for charged solute as for neutral ones, I >wonder if any one has similar experience on this issue? That's general for any solvation model. PCM/DIR seems to be better than most in this respect (see above). >5. Finally, I'd appreciate any comments on solvation effect calculations >of highly charged systems. See above, any ionic system is problematic, and anions are generally worse than cations. PCM/DIR accounts to some extent for the tailing of electron density outside the cavity, it alleviates some of the problems. There are some indications you should use smaller basis sets for the solvation calculations, at least not use diffuse functions. See, for example, T. Clark, Faraday Discuss. Chem. Soc., 1984, 78, 203-212. Regards, Per-Ola Norrby ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ * Per-Ola Norrby, Associate Professor * The Royal Danish School of Pharmacy, Dept. of Med. Chem. * Universitetsparken 2, DK 2100 Copenhagen, Denmark * tel. +45-35376777-506, +45-35370850 fax +45-35372209 * Internet: peon@medchem.dfh.dk, http://compchem.dfh.dk/ From wschnei2@ford.com Tue Oct 21 11:32:06 1997 Received: from mailfw1.ford.com for wschnei2@ford.com by www.ccl.net (8.8.3/950822.1) id KAA23995; Tue, 21 Oct 1997 10:36:16 -0400 (EDT) Received: by mailfw1.ford.com id AA18000 (InterLock SMTP Gateway 3.0 for chemistry@www.ccl.net); Tue, 21 Oct 1997 10:36:13 -0400 Message-Id: <199710211436.AA18000@mailfw1.ford.com> Received: by mailfw1.ford.com (Internal Mail Agent-1); Tue, 21 Oct 1997 10:36:13 -0400 Date: Tue, 21 Oct 1997 10:37:26 -0400 From: "William F. Schneider" To: chemistry@www.ccl.net Subject: Parallel ADF I am trying to get the parallel version of ADF-2.3 running under MPI on an IBM SP computer. The code compiles without difficulty, but when exectued gives an illegal instruction error. Has anyone else gotten the code to work on this platform? If so, can you share with me the execution environment you use, and/or any other hints? Thanks... -- +-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+ Bill Schneider e-mail: wschnei2@ford.com Chemistry Department Ford Motor Co., Drop 3083/SRL phone: (313) 323-2064 Dearborn MI 48121-2053 FAX: (313) 594-2923 From godbout@chad.scs.uiuc.edu Tue Oct 21 12:32:08 1997 Received: from chad.scs.uiuc.edu for godbout@chad.scs.uiuc.edu by www.ccl.net (8.8.3/950822.1) id LAA24468; Tue, 21 Oct 1997 11:56:47 -0400 (EDT) Received: from localhost (godbout@localhost) by chad.scs.uiuc.edu (AIX4.2/UCB 8.7/8.7) with SMTP id KAA15000; Tue, 21 Oct 1997 10:50:23 -0500 (CDT) Date: Tue, 21 Oct 1997 10:50:23 -0500 (CDT) From: Nathalie Godbout To: chemistry@www.ccl.net cc: godbout@chad.scs.uiuc.edu Subject: SUMMARY-newzmat transformation Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hello everyone, A little back I had posted a message to the list about a problem I was having with converting a pdb file with newzmat. I received several replies and I thank everyone who send their suggestions. I will look into installing Babel for the future but in the meantime, I ended up editing the pdb file. Here are the replies. Nathalie ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- Dear Nathalie: Without seeing your .pdb files, I can't quite come up with a suggestion, although I've witnessed odd behavior in newzmat before as well. You might try Babel and see if that works. It's available for any platform at the CCL ftp depot. Or send the file and I can transform it for you. Regards, -Michael ________________________________________________________________________________ Michael D. Bartberger bartberg@chem.ufl.edu TEL: (352) 392-3580 Department of Chemistry bartberg@qtp.ufl.edu FAX: (352) 846-0296 University of Florida Gainesville, FL 32611 USA ________________________________________________________________________________ ------------------------------------------------------------------------------- Hi: Babel, and I presume newzmat are choking on the lack of connectivity information. Some programs, HyperChem, etc. let you choose whether to save the connectivity. Of course for QM that should not matter, but I think that, at least for Babel, this is where the problem lies. Do you have such an option in the program you use to create the structure? Failing that (I don't know if you've done this or not and / or whether it has any bearing on your old, sucessfully-converted structures) you could just strip out the Cartesian coordinates from the .pdb file by hand. That is, for example HETATM 35 C8 NONEA 1 -2.144 -.602 2.051 1.00 .00 HETATM 36 C9 NONEA 1 -2.151 .902 -1.973 1.00 .00 HETATM 37 C2 NONEA 1 .726 -2.870 -.229 1.00 .00 ^ ^ ^ These are X Y Z this is a bit of a pain but it can be done. Good luck, -MDB ------------------------------------------------------------------------------- Natalie: Have you tried BABEL at all? JOhn -- ************************************ * John McKelvey * * Computational Science Laboratory * * Imaging Research and Development * * Building 83 * * Research Laboratories * * Eastman Kodak Company * * Rochester, NY 14650-2216 * * (V)716-477-3335 * * (F)716-722-2327 * * (E)McKelvey@Kodak.COM * ************************************ ------------------------------------------------------------------------------- Nathalie, That message indicates that newzmat it attempting to define a dihedral where three of the atoms are co-linear. If Cerius can write out cartesians in place of the PDB file you can read in straight cartesians rather than a Z-matrix. Just put a route card, title and charge/multiplicity on the front. For optimizations with redundant internal coordinates G94 will generate a linear bond for this trio, an option not defined for Z-matrices. Can you send a sample of the offending PDB file? We generalized the PDB input routines but not sufficiently to take this syntax. Douglas J. Fox Director of Technical Support help@gaussian.com ------------------------------------------------------------------------------- Do you have connect data? I think newzmat needs that. If you don't then use emacs to remove the unwanted columns. Hope this helps. Bill --- Bill King | email: Bill.King@anu.edu.au Division of Biochemistry and Molecular Biology| John Curtin School of Medical Research | phone: +61-6-279-8303 PO Box 334 Canberra ACT 2601 | FAX: +61-6-249-0415 ------------------------------------------------------------------------------- what does babel say? i use both newzmat and babel and the combination of both usually does the job even in tricky cases. hope this helps. ------------------------------------------------------------------------------- -------------------------------------------- | Nathalie Godbout | | University of Illinois at Urbana-Champaign | | 600 S. Goodwin Av. Box 23.6 | | Urbana, IL 61801 | -------------------------------------------- From daizadeh@indigo.ucdavis.edu Tue Oct 21 16:32:10 1997 Received: from indigo.ucdavis.edu for daizadeh@indigo.ucdavis.edu by www.ccl.net (8.8.3/950822.1) id PAA25945; Tue, 21 Oct 1997 15:34:12 -0400 (EDT) Received: by indigo.ucdavis.edu (940816.SGI.8.6.9/940406.SGI.AUTO) id MAA14868; Tue, 21 Oct 1997 12:36:08 -0700 From: "Iraj Daizadeh" Message-Id: <9710211235.ZM14866@indigo.ucdavis.edu> Date: Tue, 21 Oct 1997 12:35:54 -0700 X-Mailer: Z-Mail (3.2.0 26oct94 MediaMail) To: chemistry@www.ccl.net, daizadeh@indigo.ucdavis.edu Subject: Summary: Molecular Dynamics on Biological Molecules. Cc: molecular-dynamics-news@mailbase.ac.uk, katz@indigo.ucdavis.edu, antony@indigo.ucdavis.edu, stuchebr@indigo.ucdavis.edu, leee@indigo.ucdavis.edu, guo@indigo.ucdavis.edu, fink@indigo.ucdavis.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Hello. Well, it has been more than one week since the first email went out concerning software that is able to simulate classical dynamics for biomolecules on a DEC 500 au workstation. The outline of this summary is as follows: First the original question is stated. Then in chronological order all replies are presented [email addresses and other information concerning the whereabouts of the original authors (besides their names) have been deleted]. Thanks to everyone that applied and all those that are interested. Iraj. ------------------------------------------------------- Hello. Our group has just bought 2 DEC 500 au (low-end workstations) running DEC UNIX 4.0c and we are trying to find computational chemistry software packages that can calculate classical dynamics of motion for large (circa. 2000+atoms) biological molecules. It would be nice if the programs could implement a Nose/Hoover thermostat and equipped with a variety of force fields such as amber, ESFF, cff91, etc... Futhermore, the programs (which may be free or commercial) must be able to write out trajectories to some sort of archive file in ascii form for our own calculations. I will summarize the results as is customary. Thanks in advance for all contributions. Iraj. -------------------------------------------------------- 1. NAMD currently doesn't have support for DEC UNIX, and it only uses Charmm force field. NAMD runs on a variety of other platforms, though, and it should not be difficult to make a port to DEC, provided you have pvm on your machines. Otherwise, NAMD can simulate systems of large size (the largest test we did was about 500,000 atoms, and there are production simulations for systems of 32,000 atoms), maintain the temperature of the system in the simulation, and write trajectory files. VMD, on the other hand is a stand alone visualization and analysis program which can be used with a variety of file formats and can work with simulation trajectories. Sincerely, Sergei Izrailev 2) Dear Iraj, there are several programs doing the things you want. I just recall DLPOLY from Daresbury (free after signing a licence), and Amber (commercial). You'll find links to the above and several others looking throught our software page at http://antas.agraria.uniss.it go there an then click the Computational Chemistry Software link. You'll find some info also in the molmod page Hope it helps regards Bruno Manunza 3) Checkout the GROMACS code http://rugmd0.chem.rug.nl/~gmx It is definitely the fastest code on workstations for MD. Groeten, David. 4) I know that the DEC Alpha is one of the platforms supported by the academic version of CHARMM; it does have Nose/Hoover, can easily handle 30,000 atoms or more, and individual coordinate frames from a trajectory can be easily written in the CHARMM ASCII format, or in PDB format. It's not so great at supporting other force fields, for a variety of reasons (some technical). The academic version is available for a modest fee via the Karplus group at Harvard; for more info see http://yuri.harvard.edu -- Rick Venable 5) Have you tried the Amber suite of programs, or any of the Biosym software? Details for both should be easily found on the web. Clare Bostock-Smith 6) Dear Iraj, Here at Reading we use DLPOLY which is a free licensed MD software developed by Smith and Forester at Daresbury Labs in Northern England. It can do all you ask for and more :) It is written in Fortran and can be compiled for any number of architectures and requirements. I have found is immensely better than any commercially available software. I am not sure about its use outside the UK - it is written by people working for one of the research councils here in the UK and it is free for people working in the UK who sign a contract saying what they are working on - i.e. for non commercial gain etc. check out http://gserv1.dl.ac.uk/TCSC/Software/DL_POLY/main.html for more details and get back to me if I cna be of more help. Tim -- Tim Astley 7) Dear Iraj, One of the OMG software products you might be interested is AMBER. Please check the following sites for more information: http://www.oxmol.com http://www.amber.ucsf.edu/amber/amber.html Thank you for your inquiry and please let us know if we can be of help in anyway. Sincerely, NT Ton 8) 1. Since we have access to source code and since the program is written from what I can see in Python, how difficult would it be to write the dynamics trajectory snap shot by snap-shot to a file in either .pdb or .bgf form? Has this been done by someone else? Converting a trajectory files to a series of PDB files is very easy. I don't know .bgf, so I can't comment on it. Generating a trajectory directly as PDB files is more difficult, but doable. 2. What force fields does MMTK come? Would it be difficult to add new forcefields. Amber94 is implemented; others can easily be added, but there is no documentation (yet) on how to do it. 3. Can this program work on a DEC 500 au running DECK UNIX 4.0C? I see no reason why it shouldn't work, but I didn't test it on such a machine. Basically all Unix flavours should work. You should have come t the Python conference in San JOse (wjere I am right now; I have given a talk plus demonstrtions of MMTK... Konrad Hinsen -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- Iraj Daizadeh Department of Chemistry University of California One Shields Ave. Davis, CA 95616-5295 Phone: 916.754.8695 Fax: 916.752.8995 email: daizadeh@indigo.ucdavis.edu From PHogue@space.honeywell.com Wed Oct 22 14:32:19 1997 Received: from az76ms01.az76.honeywell.com for PHogue@space.honeywell.com by www.ccl.net (8.8.3/950822.1) id OAA01859; Wed, 22 Oct 1997 14:04:18 -0400 (EDT) Received: by az76ms01.az76.honeywell.com with SMTP (Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63) id <01BCDEDB.515C7870@az76ms01.az76.honeywell.com>; Wed, 22 Oct 1997 11:12:00 -0700 Message-ID: From: "Hogue, Pat (AZ76)" To: "'CCL'" Subject: Free-Ware Date: Wed, 22 Oct 1997 11:11:59 -0700 X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63 Encoding: 7 TEXT I would appreciate any advice on free software to calculate the heat of adsorption of hydrocarbon and silahydrocarbon fluids and a iron oxide surface. Thanks Pat Hogue (602) 561-3748 From mathieu@ripault.cea.fr Thu Oct 23 06:32:27 1997 Received: from nenuphar.saclay.cea.fr for mathieu@ripault.cea.fr by www.ccl.net (8.8.3/950822.1) id GAA05268; Thu, 23 Oct 1997 06:14:40 -0400 (EDT) Received: from muguet.saclay.cea.fr (muguet.saclay.cea.fr [132.166.192.6]) by nenuphar.saclay.cea.fr (8.8.5/ CEANET-ROUTER-3.0) with ESMTP id MAA14921 for ; Thu, 23 Oct 1997 12:14:30 +0200 (MET DST) Received: from cerbere.limeil.cea.fr (cerbere.bruyeres.cea.fr [132.165.76.1]) by muguet.saclay.cea.fr (8.8.5/ CEANET-ROUTER-3.0) with SMTP id MAA08656 for ; Thu, 23 Oct 1997 12:14:21 +0200 (MET DST) Received: by cerbere.limeil.cea.fr (CEANET-1.1+) id AA27074; Thu, 23 Oct 1997 12:14:07 +0200 Received: from vienne.ripault.cea.fr(132.165.32.4) by cerbere via smap id sma026584; Thu Oct 23 12:11:51 1997 Received: from vienne.ripault.cea.fr (localhost [127.0.0.1]) by ripault.cea.fr (8.8.4/8.7.3) with SMTP id JAA00937 for ; Thu, 23 Oct 1997 09:20:01 +0200 Sender: mathieu@ripault.cea.fr Message-Id: <344EFAA0.24E258B8@ripault.cea.fr> Date: Thu, 23 Oct 1997 09:20:00 +0200 From: Didier MATHIEU Organization: CEA - Le Ripault X-Mailer: Mozilla 3.0 (X11; I; Linux 2.0.27 i586) Mime-Version: 1.0 To: chemistry@www.ccl.net Subject: langage Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I am a new subscriber to CCL. I read recently a discussion about OO programming in computational chemistry, dealing mostly with Fortran, C and C++ langages. For me I have to choose a programming langage. Indeed, facing a serious need for a broad range of specific programs for molecular engineering of energetic materials (implementing simple and efficient empirical models based on molecular mechanics, groups increments, QSPR relationships...) we have these programs written by students who remain in the team for short training periods (few months). In this context of fairly long-term projects with high turnover and poor communication between succesive programmers, pieces of code written by some people must be used easily by others, hence codes must be at least modular. Experience with Fortran has shown us that training students, because of the limited time allocated, are in a hurry to implement as much as possible, and resulting code is not well structured. On the other hand, since they are mostly students in physics or chemistry, most of them do not master C or C++ , although all of them have some programming experience, for instance with Pascal or Basic. According to me, none of those free langages, C, C++ and Fortran, is a good choice and I am looking for a SIMPLE langage supporting REUSABILITY and QUICK DEVELOPMENT. Maybe some at CCL have some interesting suggestions or experience ? -- Didier MATHIEU CEA - Le Ripault, BP 16 37260 Monts (France) Tel. 33(0)2.47.34.41.85