From owner-chemistry@ccl.net Mon Dec 17 03:30:00 2007 From: "Michel Petitjean ptitjean,,itodys.jussieu.fr" To: CCL Subject: CCL: Any open-sourced generation algorithm for Connolly molecular surface?? Message-Id: <-35867-071217032506-25145-awZ5rdvj64/YFBWLUi5GWA(0)server.ccl.net> X-Original-From: Michel Petitjean Date: Mon, 17 Dec 2007 09:24:49 +0100 (MET) Sent to CCL by: Michel Petitjean [ptitjean(0)itodys.jussieu.fr] To: chemistry*|*ccl.net Subject: CCL: Re: Any open-sourced generation algorithm for Connolly molecular surface?? Do you mean: (a) Computing the surface area or (b) Viewing graphically the surface area For (a), use the freeware ASV (Monte-Carlo and/or analytical calculation): http://petitjeanmichel.free.fr/itoweb.petitjean.freeware.html#ASV The free MC routines sources are in f77, but are callable in C. It works for a null radius probe sphere (i.e. van der Waals surface or SAS). For (b), do a Google search. Remark: the analytical Connolly molecular surface algorithm fails in most situations because intersections or more than 3 spheres exist, unless atomic radii are very small (as it occurs in commercial softwares !!!). Atomic radii recommended in the literature induce intersections of 6 spheres in many cases (see Bondi, Gavezzotti, etc.). Worse: it has been demonstrated that surfaces are highly sensitive to atomic radii. ASV performs the analytical computation in all situations, for any radii values. Best regards, Michel Petitjean, DSV/iBiTec-S/SB2SM (CNRS URA 2096) CEA Saclay, bat. 528 91191 Gif-sur-Yvette Cedex Phone: +33(0)1 6908 9681 / Fax: +33(0)1 6908 4007 E-mail: petitjean*|*itodys.jussieu.fr, michel.petitjean*|*cea.fr http://petitjeanmichel.free.fr/itoweb.petitjean.spheres.html Formerly: ITODYS (CNRS, UMR 7086), 1 rue Guy de la Brosse, 75005 Paris, France. Sent to CCL by: "Jun Tie Cheng" [tjcheng**mail.sioc.ac.cn] > Hi, CCLers > > Recently I am dealing with molecular surface generation. Is there any open-sourced alogrithm for this purpose? A C/C++ library would be very appreciated. Thanks in advance. > > Best regards, > >Jay From owner-chemistry@ccl.net Mon Dec 17 04:31:01 2007 From: "Alex A. Granovsky gran|classic.chem.msu.su" To: CCL Subject: CCL: Barcelona & Harpertown Message-Id: <-35868-071217040200-8591-rNabDCjNpjqL7SYZZ8JKWQ(~)server.ccl.net> X-Original-From: "Alex A. Granovsky" Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="koi8-r" Date: Mon, 17 Dec 2007 11:28:35 +0300 MIME-Version: 1.0 Sent to CCL by: "Alex A. Granovsky" [gran^-^classic.chem.msu.su] Dear CCLers, The results of our recent benchmarks of the latest AMD's and Intel's CPUs are now publicly available at the PC GAMESS homepage at MSU, see the link in the "Performance" section. Best regards, Alex Granovsky From owner-chemistry@ccl.net Mon Dec 17 07:35:00 2007 From: "Rene Thomsen rt|molegro.com" To: CCL Subject: CCL: Molegro releases Molegro Virtual Docker 2007 v2.3.0 Message-Id: <-35869-071217073233-11350-ZDRgxfoWN6hCff7mniSEug|*|server.ccl.net> X-Original-From: "Rene Thomsen" Date: Mon, 17 Dec 2007 07:32:29 -0500 Sent to CCL by: "Rene Thomsen" [rt.(0).molegro.com] Aarhus, Denmark, December 17th, 2007 - Molegro is pleased to announce a new release of Molegro Virtual Docker, an integrated platform for predicting protein-ligand interactions available for Windows, Linux, and Mac OS X. Molegro Virtual Docker offers high-quality docking based on a novel optimization technique combined with a user interface experience focusing on usability and productivity. New features in version 2.3.0: * Ligand Energy Inspector: detailed visualization and information about all interactions. * Structural Protein Alignment: It is now possible to rigidly align proteins. * Minor user interface updates and bug fixes (see Release Notes for details). To download a trial version, please visit our company website at: http://www.molegro.com. For more information contact: Rene Thomsen, CEO Molegro Hoegh-Guldbergs Gade 10, Bldg. 1090 DK-8000 Aarhus Denmark E-mail: rt(0)molegro.com Phone: (+45) 89 42 31 65 About Molegro Molegro is a Danish company founded in 2005. Our company concentrates on developing high-performance drug discovery solutions leading to a faster drug-development process. Our goal is to provide scientifically superior products focusing on both state-of-the-art algorithms and an intuitive graphical user interface experience. From owner-chemistry@ccl.net Mon Dec 17 08:10:00 2007 From: "Oleksandr Oleksandras nablaoblada[#]yahoo.com" To: CCL Subject: CCL: NWchem visual program packages Message-Id: <-35870-071217074731-17683-qux4cyQyhyfP3wfFx46IzA(a)server.ccl.net> X-Original-From: "Oleksandr Oleksandras" Date: Mon, 17 Dec 2007 07:47:27 -0500 Sent to CCL by: "Oleksandr Oleksandras" [nablaoblada]-[yahoo.com] Hi, does anybody know any program (except WebMO) which can visualize NWchem output i.e. geometry optimization, frequencies, etc? Thank you Oleksandr From owner-chemistry@ccl.net Mon Dec 17 08:55:01 2007 From: "Tom de Greef t.f.a.d.greef|-|tue.nl" To: CCL Subject: CCL:G: PCM and TD-DFT Message-Id: <-35871-071217042823-19502-mhmGvCWKb97lUG8JeArnAA\a/server.ccl.net> X-Original-From: "Tom de Greef" Date: Mon, 17 Dec 2007 04:28:20 -0500 Sent to CCL by: "Tom de Greef" [t.f.a.d.greef|*|tue.nl] Dear all, I have a question about the use of TD-DFT calculations including PCM solvent effects using Gaussian 03. I have read in several papers that the best method to use the PCM solvent model in combination with TD-DFT is the non-equilibrium approach (when one is interested in the fast absorption processes). The solvent reaction field in the non-equilibrium regime depends in the PCM formalism on the dielectric constant at optical frequency (for CHCl3 this is 1.2). The default TD-DFT calculations in Gaussian03 assumes a non-equilibrium approach (NonEquilSolv is default). However, when I start a TD-DFT calculation in Gaussian03 for CHCl3 it automatically takes the value of the dielectric constant (for CHCl3 this is 4.9) for the calculation. Does this mean that I need to define the dielectric constant in an explicit way (by addition of the keyword: EPS=1.2)??? With best regards, Tom From owner-chemistry@ccl.net Mon Dec 17 10:17:01 2007 From: "Soren Eustis soren^_^jhu.edu" To: CCL Subject: CCL:G: Open Shell post-HF optimizations (availability in programs) Message-Id: <-35872-071217101505-10672-dEZ8JbAsKzHKf6jXauTNHA*|*server.ccl.net> X-Original-From: "Soren Eustis" Date: Mon, 17 Dec 2007 10:15:01 -0500 Sent to CCL by: "Soren Eustis" [soren[a]jhu.edu] I have been using G03W for some time now. The size of the jobs and the number I am running has convinced me that it is time to move to another platform. under win (32-bit) I am slowed by windows and limited by the memory available for use in win32. Thus, I have upgraded to a 64-bit processor system running linux. Once I did this, I was once again motivated to explore the available open-source quantum chemistry programs available. However, I have come across a problem I had forgotten about. Namely, the availability of post-HF (MPn, CCSD) geometry optimization in these packages. GAMESS does not offer this, and NWchem seems to have limited options (I have also seen in the archives that it has been noted to be relatively slow). Any suggestions for free (for academia) programs, or commercial (non-gaussian) programs capable of geometry optimization for open shell systems using post-HF methods (MPn, CCSD)? Thanks in advance. Soren Eustis From owner-chemistry@ccl.net Mon Dec 17 11:00:00 2007 From: "James Thomas Metz James.Metz*|*Abbott.com" To: CCL Subject: CCL: PBSA GBSA calculations Message-Id: <-35873-071217105144-26643-bLHr8ftGhfwSK+wGYgxfQQ[A]server.ccl.net> X-Original-From: "James Thomas Metz" Date: Mon, 17 Dec 2007 10:51:40 -0500 Sent to CCL by: "James Thomas Metz" [James.Metz*Abbott.com] PBSA / GBSA Practioners of the CCL, Typical PBSA / GBSA estimations of delta_G_binding involve a formula such as: delta_G_binding = G(PBSA)_complex - (G(PBSA)_protein + G(PBSA)_ligand)) For a set of ligands and the same protein, what is the value of estimating G(PBSA)_protein? Shouldn't this number be a constant? If not a constant, then isn't this most likely a fundamental problem of inadequate sampling? I understand that practioners may utilize the G(PBSA)_protein number to end up with estimates of delta_G_binding that are closer to experimental numbers in an absolute sense, but for the purposes of calculating a correlation coefficient, adding or subtracting even a large number makes no difference. Furthermore, if the G(PBSA)_protein numbers are not converged to the same number, then isn't this just adding noise to the estimate of delta_G_binding. Why bother adding noise if we "know" from first principles that we should be adding or subtracting a constant? I grately appreciate thoughts and comments on this issue, especially from practioners who have wrestled with these issues. Regards, Jim Metz Abbott Laboratories From owner-chemistry@ccl.net Mon Dec 17 11:34:01 2007 From: "Jeff Hammond jeff.science{:}gmail.com" To: CCL Subject: CCL:G: NWchem visual program packages Message-Id: <-35874-071217110407-3616-ING1h3kG2RvetexbP8L6YA|-|server.ccl.net> X-Original-From: "Jeff Hammond" Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=WINDOWS-1252 Date: Mon, 17 Dec 2007 08:25:07 -0600 MIME-Version: 1.0 Sent to CCL by: "Jeff Hammond" [jeff.science_._gmail.com] ECCE is FREE program with visualization capability for both quantum and classical simulations. It works with NWChem and Gaussian. Jeff Website http://ecce.pnl.gov/ Features * Support for building molecular models o Three-dimensional direct manipulation builder o Symmetry building operations and symmetry recognition o Periodic table interface with customizable atomic colors and ra= dii o Manipulation and preparation of large molecular systems for molecular dynamics calculations o Import and export of structures in a variety of standard file formats o Infinite undo capability o User created libraries of molecular structures maintained in a database o Over 7000 structures currently contained in the Structure Library including complete sets of the standard amino acids and DNA and RNA nucleotides. * Graphical user interface to a broad range of electronic structure theory types. Supported codes currently are o NWChem o Gaussian 03=99 o Gaussian 98=99 o Amica. * Other codes are registered based on user requirements. Graphical user interface for basis set selection o Over 240 basis sets are currently contained in the Basis Set Li= brary o Basis set coverage can be specified for overall structure or on an element-by-element basis o Exponents, coefficients, detailed reference and citation information is available for each basis set o Export of basis sets to several computational code file formats is possible. * Remote submission of calculations to UNIX and Linux workstations, Linux clusters, and supercomputers. Supported queue management systems are o LSF=99 o PBS=99 o NQE/NQS=99 o LoadLeveler=99 o Maui Scheduler. * Remote communications are handled using industry standard protocols o ssh/scp o telnet/ftp o rsh/rcp o Globus (developed by Argonne National Laboratory). * No installation beyond the computational code and the perl scripting lanuage is required on the remote compute servers (including no requirement for a "root" daemon). * Three-dimensional visualization and graphical display of molecular data properties while jobs are running and after completion o Vibrational frequencies o Molecular orbitals o Spin and electron densities o Dipole, quadrupole, etc. moments o Mulliken charges o Geometry optimization steps o Energies o Molecular dynamics trajectories. * Support for importing results from NWChem, Gaussian 94=99, Gaussian 98=99, and Gaussian 03=99 calculations run outside of the ECCE environment. * Extensive web based help. On Dec 17, 2007 6:47 AM, Oleksandr Oleksandras nablaoblada[#]yahoo.com wrote: > > Sent to CCL by: "Oleksandr Oleksandras" [nablaoblada]-[yahoo.com] > Hi, does anybody know any program (except WebMO) which can visualize NWch= em output i.e. geometry optimization, frequencies, etc? > Thank you > Oleksandr From owner-chemistry@ccl.net Mon Dec 17 12:12:01 2007 From: "Van Dam, HJJ \(Huub\) h.j.j.vandam+*+dl.ac.uk" To: CCL Subject: CCL: NWchem visual program packages Message-Id: <-35875-071217103001-17047-0R3agrqJ/SLf/diR5qL7hg#,#server.ccl.net> X-Original-From: "Van Dam, HJJ \(Huub\)" Content-Class: urn:content-classes:message Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset="us-ascii" Date: Mon, 17 Dec 2007 14:53:41 -0000 MIME-Version: 1.0 Sent to CCL by: "Van Dam, HJJ \(Huub\)" [h.j.j.vandam!=!dl.ac.uk] Hi Oleksandr, The obvious tool to use is ECCE (http://ecce.pnl.gov/) as this is developed by the same lab that develops NWChem for exactly that purpose. Best wishes, Huub ======================================================= Huub van Dam (h.j.j.vandam ~~ stfc.ac.uk, +44-1925-603933) ======================================================= -----Original Message----- > From: owner-chemistry ~~ ccl.net [mailto:owner-chemistry ~~ ccl.net] Sent: 17 December 2007 12:47 To: Vandam, Huub Subject: CCL: NWchem visual program packages Sent to CCL by: "Oleksandr Oleksandras" [nablaoblada]-[yahoo.com] Hi, does anybody know any program (except WebMO) which can visualize NWchem output i.e. geometry optimization, frequencies, etc? Thank you Oleksandrhttp://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txt From owner-chemistry@ccl.net Mon Dec 17 12:54:00 2007 From: "Jeff Hammond jeff.science+/-gmail.com" To: CCL Subject: CCL:G: Open Shell post-HF optimizations (availability in programs) Message-Id: <-35876-071217125130-27096-X1AMlWrp2KSLhTq+Noxoyw]-[server.ccl.net> X-Original-From: "Jeff Hammond" Content-Disposition: inline Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1 Date: Mon, 17 Dec 2007 10:48:09 -0600 MIME-Version: 1.0 Sent to CCL by: "Jeff Hammond" [jeff.science()gmail.com] Aces II (http://aces2.de) is the most thoroughly equipped package in regards to geometry optimization for high-level ab initio methods. > From their page: =================================================== Geometry optimizations using analytic gradient techniques are possible for the following schemes: HF-SCF, MBPT(2)/MP2, MBPT(3)/MP3, MBPT(4)/MP4, CC2, CCD, QCISD, CCSD, QCISD(T), CCSD(T), CCSDT-n (n=1-4), CC3, CCSDT. Except for CC3, CCSDT-n, and CCSDT (closed-shell (RHF) only), geometry optimizations are possible for RHF, UHF, ROHF, and QRHF reference functions. =================================================== The code costs nothing and is straightforward to build on a variety of 64-bit processors (Core 2 Duo, Athlon X2, Itanium, etc.). The only caveat for geometry optimization is that you must use a Z-matrix. I find Aces to be very fast until the integral files stored on disk get into the 6 GB range, at which point my modest hard drive configuration becomes the limiting factor. You're correct that NWChem is limited with respect to geometry optimization of high-level methods. It can do analytic gradients for MP2 but not coupled-cluster methods. NWChem appears slow because it is not optimized for small computers. Jeff On Dec 17, 2007 9:15 AM, Soren Eustis soren^_^jhu.edu wrote: > > Sent to CCL by: "Soren Eustis" [soren[a]jhu.edu] > I have been using G03W for some time now. The size of the jobs and the number I am running has convinced me that it is time to move to another platform. under win (32-bit) I am slowed by windows and limited by the memory available for use in win32. Thus, I have upgraded to a 64-bit processor system running linux. Once I did this, I was once again motivated to explore the available open-source quantum chemistry programs available. However, I have come across a problem I had forgotten about. Namely, the availability of post-HF (MPn, CCSD) geometry optimization in these packages. GAMESS does not offer this, and NWchem seems to have limited options (I have also seen in the archives that it has been noted to be relatively slow). > > Any suggestions for free (for academia) programs, or commercial (non-gaussian) programs capable of geometry optimization for open shell systems using post-HF methods (MPn, CCSD)? > > Thanks in advance. > > Soren Eustis> > > -- Jeff Hammond The University of Chicago From owner-chemistry@ccl.net Mon Dec 17 14:03:01 2007 From: "Elaine Meng meng|cgl.ucsf.edu" To: CCL Subject: CCL: Any open-sourced generation algorithm for Connolly molecular surface?? Message-Id: <-35877-071217131728-12294-3hD+M80JwIAeAqR5tbLDBA(-)server.ccl.net> X-Original-From: "Elaine Meng" Date: Mon, 17 Dec 2007 13:17:23 -0500 Sent to CCL by: "Elaine Meng" [meng[-]cgl.ucsf.edu] Hi Jay, The open source program "DMS" written in C can be downloaded from http://www.cgl.ucsf.edu/Overview/software.html#dms It produces a dot surface and normals, and can report surface area associated with each surface point. It can distribute the computation across multiple hosts. Best, Elaine ----- Elaine C. Meng, Ph.D. meng~!~cgl.ucsf.edu UCSF Computer Graphics Lab and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco http://www.cgl.ucsf.edu/home/meng/index.html From owner-chemistry@ccl.net Mon Dec 17 16:15:00 2007 From: "David A. Case case%a%scripps.edu" To: CCL Subject: CCL: PBSA GBSA calculations Message-Id: <-35878-071217123734-19997-hk/RS0Sx0qWVmlWrZOHTtQ:+:server.ccl.net> X-Original-From: "David A. Case" Content-Disposition: inline Content-Type: text/plain; charset=us-ascii Date: Mon, 17 Dec 2007 09:08:08 -0800 Mime-Version: 1.0 Sent to CCL by: "David A. Case" [case|scripps.edu] On Mon, Dec 17, 2007, James Thomas Metz wrote: > > PBSA / GBSA Practioners of the CCL, > > Typical PBSA / GBSA estimations of delta_G_binding involve a formula such as: > > delta_G_binding = G(PBSA)_complex - (G(PBSA)_protein + G(PBSA)_ligand)) > > For a set of ligands and the same protein, what is the value of estimating > G(PBSA)_protein? Shouldn't this number be a constant? If not a constant, > then isn't this most likely a fundamental problem of inadequate sampling? > > I understand that practioners may utilize the G(PBSA)_protein number to > end up with estimates of delta_G_binding that are closer to experimental > numbers in an absolute sense, but for the purposes of calculating a > correlation coefficient, adding or subtracting even a large number makes no > difference. > > Furthermore, if the G(PBSA)_protein numbers are not converged to the same > number, then isn't this just adding noise to the estimate of > delta_G_binding. > Why bother adding noise if we "know" from first principles that we should be > adding or subtracting a constant? > > I grately appreciate thoughts and comments on this issue, especially from > practioners who have wrestled with these issues. There is nothing wrong with your argument in any strict sense. Estimating an absolute binding energy, however, can help identify inconsistencies or missing pieces in the physical model you are using. My personal view is that we (as a field) need to learn how to compute absolute affinities reasonably well, before we can really expect relative affinities to be consistently good. And, in general, we are still a long way from such a goal. Trying to estimate absolute affinities certainly helps to keep one humble! ...dave case