From chemistry-request {*at*} ccl.net Mon Jun 27 21:11:19 2005 Received: from mercury.chem.pitt.edu (mercury.chem.pitt.edu [136.142.60.100]) by server.ccl.net (8.13.1/8.13.1) with ESMTP id j5S1BDol029012 for ; Mon, 27 Jun 2005 21:11:13 -0400 Received: from ringo.chem.pitt.edu (localhost.localdomain [127.0.0.1]) by mercury.chem.pitt.edu (Postfix) with SMTP id 36926481F for ; Mon, 27 Jun 2005 21:00:16 -0400 (EDT) Received: from 200.165.141.250 (SquirrelMail authenticated user gustavo) by mercury.chem.pitt.edu with HTTP; Mon, 27 Jun 2005 21:00:16 -0400 (EDT) Message-ID: <4971.200.165.141.250.1119920416.squirrel:at:mercury.chem.pitt.edu> Date: Mon, 27 Jun 2005 21:00:16 -0400 (EDT) Subject: SUMMARY - Software for combinatorial generation of structures From: To: X-Priority: 3 Importance: Normal X-Mailer: SquirrelMail (version 1.2.11) MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Spam-Status: No, score=0.2 required=5.0 tests=NO_REAL_NAME autolearn=no version=3.0.4 X-Spam-Checker-Version: SpamAssassin 3.0.4 (2005-06-05) on server.ccl.net Dear CCL Readers, here you have a summary of my recent request for a software for combinatorial generation of structures. Some of the answers deal with virtual synthesis and are more sophisticated than I had in mind. The kind of structures I am planning to generate are related by a common heterocyclic skeleton containing different groups (N(CH3)2, NO2, CH3, OCH3, etc) connected to it. They are not related by a chemical reaction. The conversion of the AUTOBUILD program to work under Linux would be complicated by the fact that most of the changes would involve graphical routines and I wouldn't know how to convert the routines from Silicon Graphics to Linux. I have decided to generate my structures using SMILES and then converting them to XYZ. Once again I would like to say thanks for all the answers I have received. Sincerely yours, Gustavo L.C. Moura gustavo:at:mercury.chem.pitt.edu -------------------------------------------------------------------------------- ORIGINAL MESSAGE Dear CCL Readers, I am looking for a program capable of generating structures of molecules in a combinatorial way. In other words, I want to be able to give a template and a series of fragments to the program and having the program to generate all the structures that can be obtained by connecting these fragments to the template. The program should be able to write the final structures in a format that can be converted to be read by the MOPAC program for optimization of geometry. I would like to be able to run such a program in a machine with Linux or, perhaps, Windows operating systems. Any suggestions are welcome, but I would prefer a free for academia program. I have already found in the INTERNET the AUTOBUILD program that is distributed by the INTERPROBE CHEMICAL SERVICES company (http://www.interprobe.co.uk/inter/interprobe.html). Unfortunately, the AUTOBUILD program works only on Silicon Graphics machines. Thank you very much in advance. Sincerely yours, Gustavo L.C. Moura gustavo:at:mercury.chem.pitt.edu --------------------------------------------------------------------------------- Konrad Koehler Dear Gustovao, One possible solution relying on perlmol (http://www.perlmol.org/) and SMARTS substructure notation (http://www.daylight.com/dayhtml/doc/theory/theory.smarts.html) to define reactant functional groups may be found on the following web site: http://www.perlmol.org/examples/ I would be interested in reading a summary of your responses. Best regards, Konrad --------------------------------------------------------------------------------- Armel Le Bail See a list of programs in this paper; http://journals.iucr.org/b/issues/2002/04/00/bk0108/bk0108.pdf Though they may not completely fulfill your dreams... For inorganic compounds, see: http://journals.iucr.org/j/issues/2005/02/00/cg5019/cg5019.pdf and references inside. Best wishes, Armel Le Bail ---------------------------------------------------------------------------------- Jeremy R. Greenwood Dear Gustavo, The forthcoming Schrodinger release later this year will include a new product, CombiGlide, which can do all of the above and more (with some intelligent machinery > from ligprep to cut down on knotting for example, handle pKa, stereochemistry tautomerism, etc.). Primary platform is Linux. GUI or command-line. Dicounted but not free for academic research, though it's free for teaching. Hope this helps, Jeremy ----------------------------------------------------------------------------------- John Bushnell Hi, I took a quick look at the site below, and you can freely download the source code for Autobuild. It gives all kinds of errors if I try: g77 autobuild.f on a Linux machine here, but it may be possible to clean up the code so that it will compile correctly, and/or play with compiler flags to make this work. So...if you've got somebody who is familiar with Fortran, you could probably get this to work. But it will require some time and effort. - John ------------------------------------------------------------------------------------ Tim Aitken Dear Dr Moura, I read with interest your posting on CCL net regarding combinatorial tools for library enumeration. We offer several solutions for this purpose, using either reaction based or generic core (Markush) enumeration and have software that operates on Linux, Solaris and Windows. The Combichem Enumeration Components sound like they may be the right thing for you : http://www.accelrys.com/products/accord/accordproducts/combichem_enumeration.html Please contact me if you would like any further information, or discuss specific requirements. If our solutions do not meet your needs I may be able to point you in a more appropriate direction. Best regards, Tim ------------------------------------------------------------------------------------- Gyorgy Pirok Dear Gustavo, I would like to draw your attention to a Java-based virtual synthesis program of ChemAxon called Reactor. It is fast (up to 500k reactions/hour), flexible (able to batch process reactions creating combichem libraries), and smart (evaluates reactions containing chemo-, regio- and stereoselectivity info). It supports a wide range of reaction an molecule formats, so I am sure that you can easily integrate it with MOPAC. It is also worth to mention, that all ChemAxon tools including Reactor is FREE for the academia and include API. Some useful links: Latest scientific poster about Reactor: http://www.chemaxon.com/conf/Making_Real_molecules_in_virtual_space.pdf Related user documentation: http://www.jchem.com/index.html?content=doc/user/Reactor.html You can download it from the www.chemaxon.com web site and require a FREE Academic license in an email. I hope, that you will have fun with it. Best regards, Gyvrgy