From owner-chemistry@ccl.net Tue Feb 21 00:47:00 2006 From: "Raji Raji raji++anal.chem.tohoku.ac.jp" To: CCL Subject: CCL:G: geometry convergence - Tolerance Limit in g03 Message-Id: <-30947-060221004532-7795-zAlRR4w4twQCrbh6X+RJVA-x-server.ccl.net> X-Original-From: "Raji Raji" Date: Tue, 21 Feb 2006 00:45:28 -0500 Sent to CCL by: "Raji Raji" [raji|anal.chem.tohoku.ac.jp] Revered CCL members, I am using little bigger system, it takes huge time for the optimisation in g03 (B3LYP/6-311+G(d)). Since the Value for the Maximum Force is some what closer to the Thresold, may I take the following output as converged? By the way, I am going to use that geometry for TD-DFT calculation (Not going to worry about the accuracy of the energy, little deviation is not a matter). Item Value Threshold Converged? Maximum Force 0.000695 0.000450 NO RMS Force 0.000081 0.000300 YES Maximum Displacement 0.000489 0.001800 YES RMS Displacement 0.000158 0.001200 YES Predicted change in Energy=-1.133221D-06 GradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGrad Thanks in advance. Regards, Raji. From owner-chemistry@ccl.net Tue Feb 21 01:22:01 2006 From: "David A. Case case*scripps.edu" To: CCL Subject: CCL: Help on Molecular Dynamics software Message-Id: <-30948-060219212650-21044-QUMMMr75MM7wzm02Pim1ig{:}server.ccl.net> X-Original-From: "David A. Case" Content-Disposition: inline Content-Type: text/plain; charset=us-ascii Date: Sun, 19 Feb 2006 18:26:25 -0800 Mime-Version: 1.0 Sent to CCL by: "David A. Case" [case_-_scripps.edu] On Sun, Feb 19, 2006, Ross Walker ross~!~rosswalker.co.uk wrote: > > > Does anyone know a good parallel Molecular Dynamics code that > > can handle polarizable force field such as Dang-Chang or > > AMOEBA? I am currently using TINKER, which is a very good > > package, but the code is not paralleled. > > Amber v9 which is currently scheduled for release at the end of March will > have support for parallel simulations using the Ameoba force field. Details > will be available here once the release date is firm. > (http://amber.scripps.edu) Unfortunately, Ross' post is not correct. Version 9 of Amber will be able to do serial (single-cpu) calculations with the Amoeba force field, but *not* parallel calculations. Amber (including the current version, Amber 8) does support parallel calculations for some simpler polarizable force fields. ...hope this helps....dave case From owner-chemistry@ccl.net Tue Feb 21 02:18:00 2006 From: "Sengen Sun sengensun:_:yahoo.com" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30949-060221021556-22578-Wd6HoKTgIUg3tbYrWvCE2w_-_server.ccl.net> X-Original-From: Sengen Sun Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 Date: Mon, 20 Feb 2006 23:15:48 -0800 (PST) MIME-Version: 1.0 Sent to CCL by: Sengen Sun [sengensun * yahoo.com] > From: Goedele Roos > Subject: CCL: 6-31G basis set > Date: Mon, 20 Feb 2006 19:16:38 +0100 (CET) > > -------------------------------------------------------------------------------- > > Sent to CCL by: Goedele Roos [groos]^[vub.ac.be] > Dear all, > I've read somewhere that the 6-31G basis set in > combination with HF is a bit > tricky and that one can better use a 6-31G** basis > set. Is thit correct? And > what is the reason therefore? The latter may be slightly better. But the difference is not significant because HF is lack of electron correlation. Both of them fail absurdly in prediction of regioselectivity of many cycloadditions. If you really want to improve your results, choose a method of electron correlation such as MP2 or B3LYP at least, depending on the sizes of your molecules. If your molecules are too big, you may use HF to estimate geometries, but MP2 or B3LYP are at least required to get the energies if possible as allowed by the sizes of molecules. B3LYP was once a very popular method back to 8 years ago. But B3LYP is not sensitive to the basis set after you reach 6-31G. This point was discussed by some one (maybe Houk?). My experience also told me that there is no point to use larger basis set than 6-31G in B3LYP. regards, Sengen __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From owner-chemistry@ccl.net Tue Feb 21 02:53:03 2006 From: "Mgr. Lubos Vrbka lubos.vrbka]=[uochb.cas.cz" To: CCL Subject: CCL: Help on Molecular Dynamics software Message-Id: <-30950-060221020129-17549-5ntI5CsinuADGjqzj/Ti/Q-#-server.ccl.net> X-Original-From: "Mgr. Lubos Vrbka" Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=UTF-8; format=flowed Date: Tue, 21 Feb 2006 08:01:22 +0100 MIME-Version: 1.0 Sent to CCL by: "Mgr. Lubos Vrbka" [lubos.vrbka\a/uochb.cas.cz] > Unfortunately, Ross' post is not correct. Version 9 of Amber will be able > to do serial (single-cpu) calculations with the Amoeba force field, but *not* > parallel calculations. will A9 support multipoles? regards, lubos -- ..................................................... Mgr. Lubos Vrbka Center for Biomolecules and Complex Molecular Systems Institute of Organic Chemistry and Biochemistry Academy of Sciences of the Czech Republic Prague, Czech Republic http://www.molecular.cz/~vrbka ..................................................... From owner-chemistry@ccl.net Tue Feb 21 03:28:00 2006 From: "Andreas Klamt klamt*_*cosmologic.de" To: CCL Subject: CCL:G: COSMO, CPCM, and COSMO-RS definitions Message-Id: <-30951-060221031007-24647-HYdNtXetJS2ps197Pd3RZA^server.ccl.net> X-Original-From: Andreas Klamt Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=ISO-8859-1 Date: Tue, 21 Feb 2006 09:09:59 +0100 MIME-Version: 1.0 Sent to CCL by: Andreas Klamt [klamt() cosmologic.de] Dear CCLers, this morning I again got quite angry about two papers mixing up the term COSMO, CPCM, and COSMO-RS, and therefore I like to explain them here shortly again. Please try to use them correctly and cite them correctly, when you write your papers using one of these methods. - COSMO: The conductor-like screening model is the variant of the continuum solvation models, which uses a scaled conductor boundary condition instead of the much more complicated dielectric boundary condition for the calculation of the polarization charges of a molecule in a continuum. The correct reference is [A. Klamt, G. Schüürmann, J. Chem. Soc. Perkin Trans.2 (1993) 799] COSMO has turned out to be computationally extremely efficient and robust, and to be less sensitive to outlying charge error. Therefore it has been implemented in in a wide range of quantum chemical codes. Please note that there are differences - mainly regarding the cavity construction and the radii defaults - in the various implementation, and specially between the implementations originating from my code and those originating from PCM or other codes. Nevertheless, I think all of them should be summarized under the expression COSMO. (It is important to note, that COSMO means a scaled conductor boundary condition, not a conductor itself, as often written in literature. It is not a conductor model, but a conductor-like model!) - CPCM is the implementation of COSMO in the PCM framework (V. Barone and M. Cossi, J. Phys. Chem. A 1998, 112, 1995ff), as most widely distributed and used in the Gaussian program. To be honest, I would be grateful if a citation to the original COSMO paper would be given as well when using CPCM or another re-implementation, because meanwhile 60% of the COSMO usages do not refer to COSMO anymore. - COSMO-RS (COSMO for real solvents, [1,2]) is a statistical thermodynamics post-processing of COSMO calculations, which I personally consider as the most physically sound continuum solvation model presently availably. It extends the applicability of quantum chemistry to the entire range of fluid phase thermodynamics, including mixtures and variable temperatures. COSMO-RS is based on COSMO calculations (COSMO files) for the individual solutes and solvents). COSMO-RS calculations require a COSMO-RS post-processing program, e.g. our COSMOtherm program. (Just for curiosity: I - as the inventor of COSMO - do not believe in the simple dielectric continuum solvation models at all, neither PCM nor COSMO. I am convinced that only by the RS extension a reasonable physics is put into these models. Therefore you will not find a single quantitative COSMO parameterization by me. Since 1994 I only work on COSMO-RS) 1. [1] A. Klamt, J. Phys. Chem. 99 (1995) 2224 2. [2] A. Klamt, V. Jonas, T. Buerger, J.C.W. Lohrenz, J. Phys. Chem. 102 (1998) 5074-5085 - Since Gaussian03 there is (or there was - I am not absolutely sure about the latest sub-releases) a keyword COSMORS available in PCM. This keyword has caused a lot of misuse. It was intended to trigger the right settings in PCM in order to produce COSMO files that can be used in COSMOtherm. But now many Gaussian users use this keyword and report the results as COSMO-RS calculations. This is absolutely wrong!!! These are CPCM calculations using the defaults optimized within COSMO-RS. By the way, even the term "KLAMT radii" has entered literature, since Gaussian introduced the keyword "radii=KLAMT". I am not a fan of such personalization. Please use the expression "radii optimized for COSMO-RS" or "COSMO-RS radii" and use reference [2] for these radii. Thanks for your help in avoiding more confusion Andreas Klamt P.S.: A rather complete description of COSMO and COSMO-RS is given in my recent book "COSMO-RS: From Quantum Chemistry to Fluid Phase Thermodynamcs and Drug Design", Elsevier, 2005 -- ----------------------------------------------------------------------------- Dr. habil. Andreas Klamt COSMOlogic GmbH&CoKG Burscheider Str. 515 51381 Leverkusen, Germany Tel.: +49-2171-73168-1 Fax: +49-2171-73168-9 e-mail: klamt=cosmologic.de web: www.cosmologic.de ----------------------------------------------------------------------------- COSMOlogic Your Competent Partner for Computational Chemistry and Fluid Thermodynamics ----------------------------------------------------------------------------- From owner-chemistry@ccl.net Tue Feb 21 04:02:01 2006 From: "Igor Avilov avilovi.:.averell.umh.ac.be" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30952-060221032349-893-ZeBeX5cOcO+4MGhymZI7XA^server.ccl.net> X-Original-From: "Igor Avilov" Content-class: urn:content-classes:message Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset="iso-8859-1" Date: Tue, 21 Feb 2006 09:23:29 +0100 MIME-Version: 1.0 Sent to CCL by: "Igor Avilov" [avilovi]![averell.umh.ac.be] Dear Goedele, I think that if one may not care about the polarization effects, both 6-31G and 6-31G** will give similar results. But it's not always the case. Ammonia is a notorious example, it is planar in 6-31G and pyramidal in 6-31G*. Best regards, Igor. -----Original Message----- > From: owner-chemistry[]ccl.net [mailto:owner-chemistry[]ccl.net] Sent: lundi 20 février 2006 19:22 To: Igor Avilov Subject: CCL: 6-31G basis set Sent to CCL by: Goedele Roos [groos]^[vub.ac.be] Dear all, I've read somewhere that the 6-31G basis set in combination with HF is a bit tricky and that one can better use a 6-31G** basis set. Is thit correct? And what is the reason therefore? Thanks in advance for considering my question, Best regards, goedele Drs. Goedele Roos Dienst Algemene Chemie (ALGC) Vrije Universiteit Brussel (VUB) Pleinlaan 2 B-1050 Brussels Tel: 0032-2-629 35 16 Fax: 0032-2-629 33 17http://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 Tue Feb 21 04:38:01 2006 From: "Mikael Johansson mikael.johansson]-[helsinki.fi" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30953-060221033055-3644-TvREznn3QXqVEkOlQKy07Q|*|server.ccl.net> X-Original-From: Mikael Johansson Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Date: Tue, 21 Feb 2006 10:30:50 +0200 (EET) MIME-Version: 1.0 Sent to CCL by: Mikael Johansson [mikael.johansson#helsinki.fi] Hello! On Tue, 21 Feb 2006, Sengen Sun sengensun:_:yahoo.com wrote: > ago. But B3LYP is not sensitive to the basis set after > you reach 6-31G. This point was discussed by some one > (maybe Houk?). My experience also told me that there > is no point to use larger basis set than 6-31G in > B3LYP. I disagree. Polarization functions are important also in DFT calculations. Of course, there might be some specific property that seems to be content with 6-31G. But it's true that DFT calculations don't need nearly as large basis sets as correlated wave function models, the saturarion is reached much more rapidly. Have a nice day, Mikael J. http://www.helsinki.fi/~mpjohans/ From owner-chemistry@ccl.net Tue Feb 21 05:13:00 2006 From: "Goedele Roos groos:_:vub.ac.be" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30954-060221033620-5244-cMw7IeFldX/VvJnlzQbL5w ~~ server.ccl.net> X-Original-From: Goedele Roos Date: Tue, 21 Feb 2006 09:36:11 +0100 (CET) Sent to CCL by: Goedele Roos [groos/./vub.ac.be] I have a large system (a 10-residue peptide) and I was recommended to use rather HF/3-21G than HF/6-31G, because -so was said to me- for geometry optimiations, hf/6-31G works worser than HF/3-21G, unless there were put some polarization functions on it... Is this coirrect, and why? Another point: there is no point in using larger basis set with b3LYP. But which other functional would you recommand than, with which larger basis sets are usefull? Best regards, goedele >Sent to CCL by: Sengen Sun [sengensun * yahoo.com] > >> From: Goedele Roos >> Subject: CCL: 6-31G basis set >> Date: Mon, 20 Feb 2006 19:16:38 +0100 (CET) >> >> >-------------------------------------------------------------------------------- >> >> Sent to CCL by: Goedele Roos [groos]^[vub.ac.be] >> Dear all, >> I've read somewhere that the 6-31G basis set in >> combination with HF is a bit >> tricky and that one can better use a 6-31G** basis >> set. Is thit correct? And >> what is the reason therefore? > >The latter may be slightly better. But the difference >is not significant because HF is lack of electron >correlation. Both of them fail absurdly in prediction >of regioselectivity of many cycloadditions. > >If you really want to improve your results, choose a >method of electron correlation such as MP2 or B3LYP at >least, depending on the sizes of your molecules. If >your molecules are too big, you may use HF to estimate >geometries, but MP2 or B3LYP are at least required to >get the energies if possible as allowed by the sizes >of molecules. > >B3LYP was once a very popular method back to 8 years >ago. But B3LYP is not sensitive to the basis set after >you reach 6-31G. This point was discussed by some one >(maybe Houk?). My experience also told me that there >is no point to use larger basis set than 6-31G in >B3LYP. >regards, >Sengen > > >__________________________________________________ >Do You Yahoo!? >Tired of spam? Yahoo! Mail has the best spam protection around >http://mail.yahoo.com> > > > Drs. Goedele Roos Dienst Algemene Chemie (ALGC) Vrije Universiteit Brussel (VUB) Pleinlaan 2 B-1050 Brussels Tel: 0032-2-629 35 16 Fax: 0032-2-629 33 17 From owner-chemistry@ccl.net Tue Feb 21 06:54:01 2006 From: "Mariusz Radon mariusz.radon|a|gmail.com" To: CCL Subject: CCL:G: geometry convergence - Tolerance Limit in g03 Message-Id: <-30955-060221063758-32159-pwHVnsQ1wdUfLvs+wm89Fg,server.ccl.net> X-Original-From: "Mariusz Radon" Content-Type: multipart/alternative; boundary="----=_Part_13415_23387261.1140518404401" Date: Tue, 21 Feb 2006 11:40:04 +0100 MIME-Version: 1.0 Sent to CCL by: "Mariusz Radon" [mariusz.radon{:}gmail.com] ------=_Part_13415_23387261.1140518404401 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline On 2/21/06, Raji Raji raji++anal.chem.tohoku.ac.jp wrote: > Revered CCL members, > I am using little bigger system, it takes huge time for the > optimisation in g03 (B3LYP/6-311+G(d)). Since the Value for the Maximum > Force is some what closer to the Thresold, may I take the following outpu= t > as converged? By the way, I am going to use that geometry for TD-DFT > calculation (Not going to worry about the accuracy of the energy, little > deviation is not a matter). > > Item Value Threshold Converged? > Maximum Force 0.000695 0.000450 NO > RMS Force 0.000081 0.000300 YES > Maximum Displacement 0.000489 0.001800 YES > RMS Displacement 0.000158 0.001200 YES > Predicted change in Energy=3D-1.133221D-06 > GradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGrad > > Thanks in advance. > Regards, > Raji. > Hi, Do you have oscillating or monotonically decreasing gradient? In the second case it's very likely that max. gradient'll converge in just few cycles (it's almost converged now). But if gradient is oscillating you can't be sure that more iterations help anything. Anyway, you have to check whole history of optimisation, not only the last point, to make such conclusion. If you don't take care about very accurate energies *I think* these results can be though converged, especially if gradient is oscillating. Take care, Mariusz -- Mariusz Rado\'n e-mail: mradon/at/chemia.uj.edu.pl e-mail & im: mariusz.radon/at/gmail.com ------=_Part_13415_23387261.1140518404401 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline
On 2/21/06, Raji Raji raji++anal.chem.tohoku.ac.jp <owner-chemistry . ccl.net> wrote:
Revered CCL membe= rs,
     I am using little bigger system, it takes huge time for the optimisation in g03 (B3LYP/6-311+G(d)). Since the Value for the Maximum Force is some what closer to the Thresold, may I take the following output as converged? By the way, I am going to use that geometry for TD-DFT calculation (Not going to worry about the accuracy of the energy, little deviation is not a matter).

    &= nbsp;    Item            = ;   Value     Threshold  Converged?
Maximum Force           &nbs= p;0.000695     0.000450     NO
RMS     Force           &nbs= p;0.000081     0.000300     YES
Maximum Displacement  &n= bsp;  0.000489     0.001800    = ; YES
RMS     Displacement     0.000158     0.001200     YES
P= redicted change in Energy=3D-1.133221D-06
GradGradGradGradGradGradGradG= radGradGradGradGradGradGradGradGradGradGrad

Thanks in advance.
Re= gards,
Raji.

Hi,
Do you have oscillating or monotonically decreasing gradient? In the second case it's very likely that max. gradient'll converge in just few cyc= les (it's almost converged now). But if gradient is oscillating you can't be sure that more iterations help anything. Anyway, you have to check whole history of optimisation, not only the last point, to make such conclusion.

If you don't take care about very accurate energies *I think* these results can be though converged, especially if gradient is oscillating.

Take care,
Mariusz

--
Mariusz Rado\'n
e-mail: mradon/at/chemia.uj.edu.pl
e-mail = & im: mariusz.radon/at/gmail.com
------=_Part_13415_23387261.1140518404401-- From owner-chemistry@ccl.net Tue Feb 21 07:28:01 2006 From: "Noel O Boyle no228^^^cam.ac.uk" To: CCL Subject: CCL:G: geometry convergence - Tolerance Limit in g03 Message-Id: <-30956-060221064921-464-cyRsZ4MLWnOqsFuH6qNW9w],[server.ccl.net> X-Original-From: "Noel O'Boyle" Content-Transfer-Encoding: 7bit Content-Type: text/plain Date: Tue, 21 Feb 2006 11:48:41 +0000 Mime-Version: 1.0 Sent to CCL by: "Noel O'Boyle" [no228 ~~ cam.ac.uk] Dear Raji, I cannot answer your question directly, but you may be able to get some idea of how long it will take for the geometry to converge by using GaussSum (http://gausssum.sf.net), which will plot the 'progress' of the geometry optimisation. This progress is the log of the difference between the 'Value' and the 'Threshold'. (You can do this even while the calculation is running.) Regards, Noel On Tue, 2006-02-21 at 00:57 -0500, Raji Raji raji ++anal.chem.tohoku.ac.jp wrote: > Sent to CCL by: "Raji Raji" [raji|anal.chem.tohoku.ac.jp] > > Revered CCL members, > I am using little bigger system, it takes huge time for the optimisation in g03 (B3LYP/6-311+G(d)). Since the Value for the Maximum Force is some what closer to the Thresold, may I take the following output as converged? By the way, I am going to use that geometry for TD-DFT calculation (Not going to worry about the accuracy of the energy, little deviation is not a matter). > > Item Value Threshold Converged? > Maximum Force 0.000695 0.000450 NO > RMS Force 0.000081 0.000300 YES > Maximum Displacement 0.000489 0.001800 YES > RMS Displacement 0.000158 0.001200 YES > Predicted change in Energy=-1.133221D-06 > GradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGradGrad > > Thanks in advance. > Regards, > Raji.> > > From owner-chemistry@ccl.net Tue Feb 21 08:03:00 2006 From: "Van Dam, HJJ \(Huub\) h.j.j.vandam a dl.ac.uk" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30957-060221052940-10000-/ZbNOpqCyYI5P2UCF4uNiA(~)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: Tue, 21 Feb 2006 09:48:29 -0000 MIME-Version: 1.0 Sent to CCL by: "Van Dam, HJJ \(Huub\)" [h.j.j.vandam%%dl.ac.uk] Hi Goedele, The obvious observation is that the 6-31G basis set lacks polarization functions. For sensible molecular Hartree-Fock calculations you'll probably want to include polarization functions that help describe the deformation of the atomic densities to describe the bonds. If I remember correctly there are molecules that are described so poorly if no polarization functions are used that the atoms become unbound. I suppose this is what you mean by "tricky". The 6-31G* basis set gives you polarization functions for elements other than hydrogen, 6-31G** is essentially the same as 6-31G* but also has polarization functions for hydrogen. I hope this helps... Best wishes, Huub ========================================================== Huub van Dam (h.j.j.vandam|,|dl.ac.uk, +44-1925-603933) ========================================================== -----Original Message----- > From: owner-chemistry|,|ccl.net [mailto:owner-chemistry|,|ccl.net] Sent: 20 February 2006 18:19 To: Vandam, Huub Subject: CCL: 6-31G basis set Sent to CCL by: Goedele Roos [groos]^[vub.ac.be] Dear all, I've read somewhere that the 6-31G basis set in combination with HF is a bit tricky and that one can better use a 6-31G** basis set. Is thit correct? And what is the reason therefore? Thanks in advance for considering my question, Best regards, goedele Drs. Goedele Roos Dienst Algemene Chemie (ALGC) Vrije Universiteit Brussel (VUB) Pleinlaan 2 B-1050 Brussels Tel: 0032-2-629 35 16 Fax: 0032-2-629 33 17http://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 Tue Feb 21 09:03:00 2006 From: "Goedele Roos groos[]vub.ac.be" To: CCL Subject: CCL:G: a correct input for AMBER Message-Id: <-30958-060221085622-2373-O22h6vxy8zCQ9miA47XvhA/./server.ccl.net> X-Original-From: Goedele Roos Date: Tue, 21 Feb 2006 14:56:13 +0100 (CET) Sent to CCL by: Goedele Roos [groos=vub.ac.be] Dear All, Can someone helps me to generate a correct Amber-input file? I tried to do it with Gaussian 03, but...my calculation crashes with "MM function not complete". I tried to do it by hand (I 've checked every atom type) but the same error occurs. Who can help me? Thak you in advance! Goedele Drs. Goedele Roos Dienst Algemene Chemie (ALGC) Vrije Universiteit Brussel (VUB) Pleinlaan 2 B-1050 Brussels Tel: 0032-2-629 35 16 Fax: 0032-2-629 33 17 From owner-chemistry@ccl.net Tue Feb 21 11:01:00 2006 From: "Green Power powergreen-*-gmail.com" To: CCL Subject: CCL:G: Molekel Message-Id: <-30959-060217002306-19461-l4spswK+vrVOmvLGAO5J1Q]|[server.ccl.net> X-Original-From: Green Power Content-Type: multipart/alternative; boundary="----=_Part_14582_24943892.1140150426225" Date: Thu, 16 Feb 2006 23:27:06 -0500 MIME-Version: 1.0 Sent to CCL by: Green Power [powergreen(~)gmail.com] ------=_Part_14582_24943892.1140150426225 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Hi, Anybody know if MOLEKEL on Win9x support g03 output files? It seems that it can not be used to calculate orbitals with g03 out files due to the problem of basie-stet. Thanks ------=_Part_14582_24943892.1140150426225 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline
Hi,
 
Anybody know if MOLEKEL on Win9x support g03 output files? It see= ms that it can not be used to calculate orbitals with g03 out files du= e to the problem of basie-stet.
Thanks
------=_Part_14582_24943892.1140150426225-- From owner-chemistry@ccl.net Tue Feb 21 11:36:00 2006 From: "Andrew T Pudzianowski andrew.pudzianowski]![bms.com" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30960-060221110144-705-F6yy1Oo6HZqemkYqeAG98g _ server.ccl.net> X-Original-From: Andrew T Pudzianowski Content-return: prohibited Content-type: multipart/alternative; boundary="Boundary_(ID_/TAZBcwu9u4/CNskFl4qxQ)" Date: Tue, 21 Feb 2006 09:54:44 -0500 MIME-version: 1.0 Sent to CCL by: Andrew T Pudzianowski [andrew.pudzianowski- -bms.com] This is a multi-part message in MIME format. --Boundary_(ID_/TAZBcwu9u4/CNskFl4qxQ) Content-type: text/plain; charset=ISO-8859-1; format=flowed Content-transfer-encoding: 7BIT Hello. In general, it's not true that B3LYP results don't improve with basis sets beyond 6-31G. Certainly for studying hydrogen-bonded complexes, B3LYP parallels MP2 rather well, in that the 6-31+G(d,p) basis set does an excellent job on optimized geometries but energies are more dependable with 6-311+G(d,p) or even higher levels of polarization functions. Note especially that diffuse functions are absolutely necessary. You might want to check the following and references therein: A. Pudzianowski, J. Phys. Chem. 100, 4781-4789 (1996). Regards....Andrew Sengen Sun sengensun:_:yahoo.com wrote: >Sent to CCL by: Sengen Sun [sengensun * yahoo.com] > > > >>From: Goedele Roos >>Subject: CCL: 6-31G basis set >>Date: Mon, 20 Feb 2006 19:16:38 +0100 (CET) >> >> >> >> >-------------------------------------------------------------------------------- > > >> Sent to CCL by: Goedele Roos [groos]^[vub.ac.be] >> Dear all, >> I've read somewhere that the 6-31G basis set in >>combination with HF is a bit >> tricky and that one can better use a 6-31G** basis >>set. Is thit correct? And >> what is the reason therefore? >> >> > >The latter may be slightly better. But the difference >is not significant because HF is lack of electron >correlation. Both of them fail absurdly in prediction >of regioselectivity of many cycloadditions. > >If you really want to improve your results, choose a >method of electron correlation such as MP2 or B3LYP at >least, depending on the sizes of your molecules. If >your molecules are too big, you may use HF to estimate >geometries, but MP2 or B3LYP are at least required to >get the energies if possible as allowed by the sizes >of molecules. > >B3LYP was once a very popular method back to 8 years >ago. But B3LYP is not sensitive to the basis set after >you reach 6-31G. This point was discussed by some one >(maybe Houk?). My experience also told me that there >is no point to use larger basis set than 6-31G in >B3LYP. >regards, >Sengen > > >__________________________________________________ >Do You Yahoo!? >Tired of spam? Yahoo! Mail has the best spam protection around >http://mail.yahoo.com> > > > > -- ---------------------------------------------------------------------------------------------- Andrew T. Pudzianowski, Ph.D. Computer-Aided Drug Design Bristol-Myers Squibb PRI Box 4000 Princeton NJ 08543-4000 office: (609) 252-4248 fax : (609) 252-6030 ))))))))))))))))))))))))))))))))))))))))))))) I used to shave with Ockham's razor but I kept getting Dedekind cuts. ((((((((((((((((((((((((((((((((((((((((((((( --Boundary_(ID_/TAZBcwu9u4/CNskFl4qxQ) Content-type: text/html; charset=ISO-8859-1 Content-transfer-encoding: 7BIT Hello. In general, it's not true that B3LYP results don't improve with basis sets beyond 6-31G. Certainly for studying hydrogen-bonded complexes, B3LYP parallels MP2 rather well, in that the 6-31+G(d,p) basis set does an excellent job on optimized geometries but energies are more dependable with 6-311+G(d,p) or even higher levels of polarization functions. Note especially that diffuse functions are absolutely necessary. You might want to check the following and references therein: A. Pudzianowski, J. Phys. Chem. 100, 4781-4789 (1996).

Regards....Andrew

Sengen Sun sengensun:_:yahoo.com wrote: 
Sent to CCL by: Sengen Sun [sengensun * yahoo.com]

  
From: Goedele Roos <groos|a|vub.ac.be> 
Subject: CCL: 6-31G basis set 
Date: Mon, 20 Feb 2006 19:16:38 +0100 (CET) 


    
--------------------------------------------------------------------------------
  
 Sent to CCL by: Goedele Roos [groos]^[vub.ac.be]
 Dear all,
 I've read somewhere that the 6-31G basis set in
combination with HF is a bit
 tricky and that one can better use a 6-31G** basis
set. Is thit correct? And
 what is the reason therefore?
    

The latter may be slightly better. But the difference
is not significant because HF is lack of electron
correlation. Both of them fail absurdly in prediction
of regioselectivity of many cycloadditions. 

If you really want to improve your results, choose a
method of electron correlation such as MP2 or B3LYP at
least, depending on the sizes of your molecules. If
your molecules are too big, you may use HF to estimate
geometries, but MP2 or B3LYP are at least required to
get the energies if possible as allowed by the sizes
of molecules. 

B3LYP was once a very popular method back to 8 years
ago. But B3LYP is not sensitive to the basis set after
you reach 6-31G. This point was discussed by some one
(maybe Houk?). My experience also told me that there
is no point to use larger basis set than 6-31G in
B3LYP.
regards,
Sengen


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.comCHEMISTRY]^[ccl.net or use:
      http://www.ccl.net/cgi-bin/ccl/send_ccl_message

E-mail to administrators: CHEMISTRY-REQUEST]^[ccl.net or use
      http://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp://www.ccl.net/chemistry/sub_unsub.shtml

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.net/jobs 
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/htdig  (login: ccl, Password: search)http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/


-- 
----------------------------------------------------------------------------------------------
Andrew T. Pudzianowski, Ph.D.  
Computer-Aided Drug Design     
Bristol-Myers Squibb PRI     
Box 4000
Princeton NJ 08543-4000
office: (609) 252-4248
fax   : (609) 252-6030
)))))))))))))))))))))))))))))))))))))))))))))

I used to shave with Ockham's razor
but I kept getting Dedekind cuts.

(((((((((((((((((((((((((((((((((((((((((((((
--Boundary_(ID_/TAZBcwu9u4/CNskFl4qxQ)-- From owner-chemistry@ccl.net Tue Feb 21 12:11:01 2006 From: "Debasis Sengupta dxs ~ cfdrc.com" To: CCL Subject: CCL:G: Molekel Message-Id: <-30961-060221112102-12754-nIjecOVyu92Qw9gYdighwg=server.ccl.net> X-Original-From: Debasis Sengupta Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii; format=flowed Date: Tue, 21 Feb 2006 10:22:33 -0600 MIME-Version: 1.0 Sent to CCL by: Debasis Sengupta [dxs::cfdrc.com] All you have to is to change the "Gaussian 03" to "Gaussian 98" in the following section of the output. Hope this helps. ...Debasis ****************************************** Gaussian 03: IA32W-G03RevC.02 12-Jun-2004 21-Feb-2006 ****************************************** Green Power powergreen-*-gmail.com wrote: >Sent to CCL by: Green Power [powergreen(~)gmail.com] > > >------=_Part_14582_24943892.1140150426225 >Content-Type: text/plain; charset=ISO-8859-1 >Content-Transfer-Encoding: quoted-printable >Content-Disposition: inline > >Hi, > >Anybody know if MOLEKEL on Win9x support g03 output files? It seems that it >can not be used to calculate orbitals with g03 out files due to the problem >of basie-stet. >Thanks > >------=_Part_14582_24943892.1140150426225 >Content-Type: text/html; charset=ISO-8859-1 >Content-Transfer-Encoding: quoted-printable >Content-Disposition: inline > >
Hi,
>
 
>
Anybody know if MOLEKEL on Win9x support g03 output files? It see= >ms that it can not be used to calculate orbitals with g03 out files du= >e to the problem of basie-stet.
>
Thanks
> >------=_Part_14582_24943892.1140150426225--> > > > > From owner-chemistry@ccl.net Tue Feb 21 12:46:00 2006 From: "angelo vargas vargas*chem.ethz.ch" To: CCL Subject: CCL:G: Molekel Message-Id: <-30962-060221123924-25388-d+AWwi+JC4K7+KPGADzFbQ!A!server.ccl.net> X-Original-From: angelo vargas Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Tue, 21 Feb 2006 18:03:45 +0100 MIME-Version: 1.0 Sent to CCL by: angelo vargas [vargas]-[chem.ethz.ch] hi Green Power you should check the latest version of Molekel (4.6) http://www.cscs.ch/a-display.php?id=138 it is still work in progress, but 4.6 should handle G03 output files. Angelo _______________________________________________________________________ Dr. Angelo Vargas Swiss Federal Institute of Technology (ETHZ) Institute for Chemical and Bio Engineering ETH Hönggerberg Room HCl E 129 Tel.: 0041/44/633 42 32 Fax : 0041/44/632 11 63 Zürich - Switzerland E-mail: vargas_-_chem.ethz.ch ________________________________________________________________________ From owner-chemistry@ccl.net Tue Feb 21 13:21:01 2006 From: "Pablo Vitoria pablo.vitoria===ehu.es" To: CCL Subject: CCL:G: Molekel Message-Id: <-30963-060221123550-23507-ZCu3eBnH8A98O2NDtLFodA()server.ccl.net> X-Original-From: "Pablo Vitoria" Content-Transfer-Encoding: 7bit Content-Type: text/plain; format=flowed; charset="iso-8859-1"; reply-type=original Date: Tue, 21 Feb 2006 17:51:28 +0100 MIME-Version: 1.0 Sent to CCL by: "Pablo Vitoria" [pablo.vitoria*ehu.es] Hi, You should (slightly) change the G03 output as in http://educ.gaussian.com/visual/Struc/html/StrucMolekel.htm and there should be no problem. Best regards Pablo ----- Original Message ----- > From: "Green Power powergreen-*-gmail.com" To: "Vitoria, Pablo " Sent: Tuesday, February 21, 2006 5:03 PM Subject: CCL:G: Molekel > Sent to CCL by: Green Power [powergreen(~)gmail.com] > Hi, > > Anybody know if MOLEKEL on Win9x support g03 output files? It seems that > it > can not be used to calculate orbitals with g03 out files due to the > problem > of basie-stet. > Thanks From owner-chemistry@ccl.net Tue Feb 21 13:56:01 2006 From: "Jozsef Csontos jozsefcsontos^^^creighton.edu" To: CCL Subject: CCL: PES vs. # of CPUs Message-Id: <-30964-060221123452-23208-3QStjJbQOz83MthAqNW8yg=server.ccl.net> X-Original-From: Jozsef Csontos Content-Type: multipart/mixed; boundary="=-zyytbE2j78JtGWjlqaww" Date: Tue, 21 Feb 2006 11:34:37 -0600 Mime-Version: 1.0 Sent to CCL by: Jozsef Csontos [jozsefcsontos-*-creighton.edu] --=-zyytbE2j78JtGWjlqaww Content-Type: text/plain Content-Transfer-Encoding: 7bit Hi Joop, first of all, list members, thanks for the useful answers. Secondly, I don't want to denigrate the company, so I only attached the cartesian input. (It was calculated in water without the first shell correction term in solvation energy using the pwpw91/6-31g* method with 0.04 trust radius. But this is not so important, because since then, I have calculated it using the default options and I also got processor dependency on the results.) Then it was computed in gas-phase and I got the same answer independently from the computer configuration. Well, my opinion is that I have the right MPI library and because I got the QM package in binary the source of the problem is the solvation model implementation. Interestingly (I could use bolder words), in some cases, I got different results when I just repeated the calculations on the same configuration. (starting-time dependence? :)) The other side is that I'm getting tired to test their packages, this is the second major version where I have found bugs and frankly I didn't search for those, I just wanted to use the programs. In both cases, the solution/answer was to buy the next version. In my opinion this is unacceptable. Well, if you use the latest, which was announced two or three months ago you might be out of trouble, but nobody expects the Spanish Inquisition. Best wishes, Jozsef On Thu, 2006-02-09 at 02:18 -0500, Joop van lenthe joop|a|chem.uu.nl wrote: > Sent to CCL by: Joop van lenthe [joop^-^chem.uu.nl] > Hi Jozsef could you send the data you used as input > I would like to try to make it less theoretical > Joop > On 1-feb-06, at 23:41, Jozsef Csontos jozsefcsontos~~creighton.edu > wrote: > > > Sent to CCL by: Jozsef Csontos [jozsefcsontos]|[creighton.edu] > > Hi, > > > > my short question is: > > is it possible to get different optimized final structures using the > > same input and architecture, OS, software (and so on), but different > > number of processors? > > > > My longer comments are: > > I understand that we use approximations (in QM and in life too:) and > > there might be error accumulations and our number representation > > depends > > on the hardware etc. But, I'm quite sure the answer is no. Maybe I'm > > wrong, so I tell you the story. > > > > The story: > > I have been using a commercial quantum package, which uses the MPI > > libraries (MPICH) for parallelization. I was about to map the PES of > > several tripeptides in water using DFT. (Of course, the molecules are > > flexible and the PESs are flat and more more approximations) I'm not > > totally surprised that the results are different on 1 node with 1 > > processor and 6 node with 6 processors. (In the first case there is no > > MPI and in the second one there is.) Interestingly, the results are > > different on 3 nodes with 6 processors (2CPUs/node) and on 6 nodes with > > 6 processors (1CPU/node), in addition the calculation using the > > previous > > 6 processor configuration (2CPUs/node) differs from the 4 CPUs config > > (2CPUs/node) ones. > > The energy differences between the structures are small about > > 1kcal/mol, > > however, the rmsd is quite large 1.9 Angstrom considering heavy atoms > > and one of these structures is the deepest minimum on the PES, so these > > structures belong to the low energy region. I haven't checked all the > > structures whether they are minima or other stationary points. I guess > > this is not related to the question, but if all of them were minima my > > question would be louder. > > > > I'm wondering what your opinions are? > > > > Best wishes, > > > > Jozsef > > > > -- > > Jozsef Csontos, Ph.D. > > > > Department of Biomedical Sciences > > Creighton University, > > Omaha, NE> To recover the email address of the author of the message, please > > change> > > Search Messages: http://www.ccl.net/htdig (login: ccl, Password: > > search)> > > -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+- > > +-+ > > > > > > > =============================== > Joop van Lenthe > Theoretical Chemistry Group Utrecht > tel. -31-30-2532733 > joop[#]chem.uu.nl > ===============================> > > -- Jozsef Csontos, Ph.D. Department of Biomedical Sciences Creighton University, Omaha, NE --=-zyytbE2j78JtGWjlqaww Content-Disposition: attachment; filename=fgg.xyz Content-Type: text/plain; name=fgg.xyz; charset=utf-8 Content-Transfer-Encoding: 7bit C1 -4.743000 8.126000 -3.239000 C2 -5.986000 8.450000 -2.424000 O3 -5.889000 9.042000 -1.349000 H4 -4.538000 7.056000 -3.207000 H5 -3.874000 8.652000 -2.844000 H6 -4.876000 8.424000 -4.279000 N7 -7.156000 8.053000 -2.932000 C8 -8.470000 8.256000 -2.311000 C9 -8.828000 7.137000 -1.307000 O10 -9.874000 7.197000 -0.658000 C11 -9.534000 8.298000 -3.437000 C12 -9.652000 7.060000 -4.319000 C13 -8.841000 6.923000 -5.466000 C14 -10.592000 6.053000 -4.008000 C15 -8.960000 5.785000 -6.286000 C16 -10.714000 4.918000 -4.831000 C17 -9.897000 4.783000 -5.970000 H18 -7.155000 7.571000 -3.820000 H19 -8.486000 9.208000 -1.777000 H20 -9.334000 9.156000 -4.080000 H21 -10.512000 8.503000 -2.999000 H22 -8.127000 7.692000 -5.727000 H23 -11.224000 6.148000 -3.135000 H24 -8.334000 5.681000 -7.161000 H25 -11.433000 4.148000 -4.588000 H26 -9.988000 3.909000 -6.601000 N27 -7.967000 6.122000 -1.178000 C28 -8.129000 4.936000 -0.339000 C29 -8.080000 3.642000 -1.160000 O30 -7.811000 2.575000 -0.607000 H31 -7.118000 6.187000 -1.720000 H32 -7.326000 4.919000 0.399000 H33 -9.066000 4.955000 0.219000 N34 -8.318000 3.733000 -2.473000 C35 -8.252000 2.636000 -3.440000 C36 -7.222000 2.888000 -4.549000 O37 -7.152000 2.119000 -5.509000 H38 -8.581000 4.640000 -2.834000 H39 -8.016000 1.684000 -2.962000 H40 -9.234000 2.516000 -3.898000 N41 -6.426000 3.958000 -4.432000 C42 -5.429000 4.369000 -5.411000 H43 -6.532000 4.515000 -3.597000 H44 -4.671000 3.595000 -5.543000 H45 -4.932000 5.281000 -5.083000 H46 -5.894000 4.565000 -6.378000 --=-zyytbE2j78JtGWjlqaww-- From owner-chemistry@ccl.net Tue Feb 21 14:31:00 2006 From: "Jozsef Csontos jozsefcsontos]~[creighton.edu" To: CCL Subject: CCL:G: Molekel Message-Id: <-30965-060221124258-26542-HqQUl+yJT/FSnh95K3d6SQ],[server.ccl.net> X-Original-From: Jozsef Csontos Content-Transfer-Encoding: 7bit Content-Type: text/plain Date: Tue, 21 Feb 2006 11:42:54 -0600 Mime-Version: 1.0 Sent to CCL by: Jozsef Csontos [jozsefcsontos*o*creighton.edu] Hi, have you tried changing the lines in beginning of the output? > from: ******************* Gaussian G03: ***************** to: ******************* Gaussian G98: ***************** Jozsef On Tue, 2006-02-21 at 11:25 -0500, Green Power powergreen-*-gmail.com wrote: > Sent to CCL by: Green Power [powergreen(~)gmail.com] > > > ------=_Part_14582_24943892.1140150426225 > Content-Type: text/plain; charset=ISO-8859-1 > Content-Transfer-Encoding: quoted-printable > Content-Disposition: inline > > Hi, > > Anybody know if MOLEKEL on Win9x support g03 output files? It seems that it > can not be used to calculate orbitals with g03 out files due to the problem > of basie-stet. > Thanks > > ------=_Part_14582_24943892.1140150426225 > Content-Type: text/html; charset=ISO-8859-1 > Content-Transfer-Encoding: quoted-printable > Content-Disposition: inline > >
Hi,
>
 
>
Anybody know if MOLEKEL on Win9x support g03 output files? It see= > ms that it can not be used to calculate orbitals with g03 out files du= > e to the problem of basie-stet.
>
Thanks
> > ------=_Part_14582_24943892.1140150426225--> > > -- Jozsef Csontos, Ph.D. Department of Biomedical Sciences Creighton University, Omaha, NE From owner-chemistry@ccl.net Tue Feb 21 16:31:00 2006 From: "Christopher Cramer cramer{:}chem.umn.edu" To: CCL Subject: CCL:G: 6-31G basis set Message-Id: <-30966-060221093634-23362-JNfsBtS0GKXNRP8Fscb4Qg*o*server.ccl.net> X-Original-From: Christopher Cramer Content-Type: multipart/alternative; boundary=Apple-Mail-2--986972513 Date: Tue, 21 Feb 2006 07:36:24 -0600 Mime-Version: 1.0 (Apple Message framework v746.2) Sent to CCL by: Christopher Cramer [cramer(!)chem.umn.edu] --Apple-Mail-2--986972513 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed > Sent to CCL by: Goedele Roos [groos/./vub.ac.be] > I have a large system (a 10-residue peptide) and I was recommended > to use rather HF/3-21G than HF/6-31G, because -so was said to me- > for geometry optimiations, hf/6-31G works worser than HF/3-21G, > unless there were put some polarization functions on it... > Is this coirrect, and why? > Don Truhlar and I developed the MIDI! basis set some years ago in part to permit efficient geometry optimization of molecules like that described below. We found that polarization functions on heteroatoms are too important to sacrifice, but those on C (and H) are not needed in order to obtain good geometries and charge distributions. We optimized the d exponents on the heteroatoms within the context of this choice. The basis set is available at the PNL website and is also coded into the popular Gaussian suite of programs (where it is called midix). The original MIDI! paper was: The MIDI! Basis Set for Quantum Mechanical Calculations of Molecular Geometries and Partial Charges Easton, R. E.; Giesen, D. J.; Welch, A.; Cramer, C. J.; Truhlar, D. G. Theor. Chim. Acta 1996, 93, 281. We present a series of calculations designed to identify an economical basis set for geometry optimizations and partial charge calculations on medium-size molecules, including neutrals, cations, and anions, with special emphasis on functional groups that are important for biomolecules and drug design. A new combination of valence basis functions and polarization functions, called the MIDI! basis set, is identified as a good compromise of speed and accuracy, yielding excellent geometries and charge balances at a cost that is as affordable as possible for large molecules. The basis set is optimized for molecules containing H, C, N, O, F, P, S, and Cl. Although much smaller than the popular 6-31G* basis set, in direct comparisons it yields more accurate geometries and charges as judged compared to MP2/cc-pVDZ calculations. Chris Cramer On Feb 21, 2006, at 4:32 AM, Goedele Roos groos:_:vub.ac.be wrote: > Another point: there is no point in using larger basis set with > b3LYP. But which other functional would you recommand than, with > which larger basis sets are usefull? > Best regards, goedele > > >> Sent to CCL by: Sengen Sun [sengensun * yahoo.com] >> >>> From: Goedele Roos >>> Subject: CCL: 6-31G basis set >>> Date: Mon, 20 Feb 2006 19:16:38 +0100 (CET) >>> >>> >> --------------------------------------------------------------------- >> ----------- >>> >>> Sent to CCL by: Goedele Roos [groos]^[vub.ac.be] >>> Dear all, >>> I've read somewhere that the 6-31G basis set in >>> combination with HF is a bit >>> tricky and that one can better use a 6-31G** basis >>> set. Is thit correct? And >>> what is the reason therefore? >> >> The latter may be slightly better. But the difference >> is not significant because HF is lack of electron >> correlation. Both of them fail absurdly in prediction >> of regioselectivity of many cycloadditions. >> >> If you really want to improve your results, choose a >> method of electron correlation such as MP2 or B3LYP at >> least, depending on the sizes of your molecules. If >> your molecules are too big, you may use HF to estimate >> geometries, but MP2 or B3LYP are at least required to >> get the energies if possible as allowed by the sizes >> of molecules. >> >> B3LYP was once a very popular method back to 8 years >> ago. But B3LYP is not sensitive to the basis set after >> you reach 6-31G. This point was discussed by some one >> (maybe Houk?). My experience also told me that there >> is no point to use larger basis set than 6-31G in >> B3LYP. >> regards, >> Sengen >> >> >> __________________________________________________ >> Do You Yahoo!? >> Tired of spam? Yahoo! Mail has the best spam protection around >> http://mail.yahoo.com> >> >> >> > > Drs. Goedele Roos > Dienst Algemene Chemie (ALGC) > Vrije Universiteit Brussel (VUB) > Pleinlaan 2 > B-1050 Brussels > Tel: 0032-2-629 35 16 > Fax: 0032-2-629 33 17 > > > > -= This is automatically added to each message by the mailing > script =- > To recover the email address of the author of the message, please > change> Conferences: http://server.ccl.net/chemistry/announcements/ > conferences/ > > Search Messages: http://www.ccl.net/htdig (login: ccl, Password: > search)> > -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+- > +-+-+ > > > -- Christopher J. Cramer University of Minnesota Department of Chemistry 207 Pleasant St. SE Minneapolis, MN 55455-0431 -------------------------- Phone: (612) 624-0859 || FAX: (612) 626-2006 Mobile: (952) 297-2575 cramer . pollux.chem.umn.edu http://pollux.chem.umn.edu/~cramer (website includes information about the textbook "Essentials of Computational Chemistry: Theories and Models, 2nd Edition") --Apple-Mail-2--986972513 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1
Sent to CCL by: Goedele Roos = [groos/./vub.ac.be]
I have a large system (a = 10-residue peptide) and I was recommended to use rather HF/3-21G than = HF/6-31G, because -so was said to me- for geometry optimiations, = hf/6-31G works worser than HF/3-21G, unless there were put some = polarization functions on it...
Is this = coirrect, and why?


Don Truhlar and I developed the = MIDI! basis set some years ago in part to permit efficient geometry = optimization of molecules like that described below. We found that = polarization functions on heteroatoms are too important to sacrifice, = but those on C (and H) are not needed in order to obtain good geometries = and charge distributions. We optimized the d exponents on the = heteroatoms within the context of this choice. The basis set is = available at the PNL website and is also coded into the popular Gaussian = suite of programs (where it is called midix). The original MIDI! paper = was:

The MIDI! Basis Set for Quantum = Mechanical Calculations of Molecular Geometries and Partial = Charges

Easton, R. E.; Giesen, D. J.; Welch, A.; Cramer, C. J.; Truhlar, = D. G.

Theor. Chim. Acta = 1996, 93, 281.

We present a series of calculations designed to identify an = economical basis set for geometry optimizations and partial charge = calculations on medium-size molecules, including neutrals, cations, and = anions, with special emphasis on functional groups that are important = for biomolecules and drug design. A new combination of valence basis = functions and polarization functions, called the MIDI! basis set, is = identified as a good compromise of speed and accuracy, yielding = excellent geometries and charge balances at a cost that is as affordable = as possible for large molecules. The basis set is optimized for = molecules containing H, C, N, O, F, P, S, and Cl. Although much smaller = than the popular 6-31G* basis set, in direct comparisons it yields more = accurate geometries and charges as judged compared to MP2/cc-pVDZ = calculations.

Chris Cramer


On Feb 21, = 2006, at 4:32 AM, Goedele Roos groos:_:vub.ac.be wrote:

Another point: there is no point in using larger = basis set with b3LYP. But which other functional would you recommand = than, with which larger basis sets are usefull?
Best regards, goedele=A0


Sent to = CCL by: Sengen Sun [sengensun * yahoo.com]

From: Goedele Roos <groos|a|vub.ac.be>=A0
Subject: = CCL: 6-31G basis set=A0
Date: = Mon, 20 Feb 2006 19:16:38 +0100 (CET)=A0



=A0Sent to CCL by: Goedele Roos = [groos]^[vub.ac.be]
=A0Dear all,
=A0I've = read somewhere that the 6-31G basis set in
=A0tricky and that one can better = use a 6-31G** basis
set. Is thit correct? = And
=A0what is the reason = therefore?

The latter may be slightly = better. But the difference
is not = significant because HF is lack of electron
of regioselectivity of many cycloadditions.=A0

If you really = want to improve your results, choose a
method = of electron correlation such as MP2 or B3LYP at
least, depending on the sizes of your molecules. = If
your molecules are too big, you may use HF to = estimate
geometries, but MP2 or B3LYP are = at least required to
get the energies if = possible as allowed by the sizes
of = molecules.=A0

B3LYP = was once a very popular method back to 8 years
ago. But B3LYP is not sensitive to the basis set = after
you reach 6-31G. This point was = discussed by some one
(maybe Houk?). My = experience also told me that there
is no point = to use larger basis set than 6-31G in
regards,
Sengen


Do You Yahoo!?
Tired of = spam?=A0 Yahoo! Mail has = the best spam protection around=A0




Drs. Goedele Roos
Dienst Algemene Chemie (ALGC)
Vrije Universiteit Brussel (VUB)
Pleinlaan 2
B-1050 = Brussels
Tel: 0032-2-629 35 16
Fax: 0032-2-629 33 17



-=3D This is automatically added to each message by = the mailing script =3D-
To recover = the email address of the author of the message, please change
the strange characters on the top line to the . = sign. You can also
look up the = X-Original-From: line in the mail header.

E-mail to = subscribers: CHEMISTRY . ccl.net = or use:

E-mail to administrators: CHEMISTRY-REQUEST . ccl.net = or use

Subscribe/Unsubscribe:=A0

Before posting, check wait time = at: http://www.ccl.net

http:/= /server.ccl.net/chemistry/announcements/conferences/

Search = Messages: http://www.ccl.net/htdig=A0 (login: ccl, Password: = search)

If your mail bounces from CCL with 5.7.1 error, = check:






--


Christopher J. Cramer

University of Minnesota

Department of Chemistry

207 Pleasant St. = SE

Minneapolis, MN 55455-0431

--------------------------

=

Phone:=A0 (612) 624-0859 || FAX:=A0 (612) 626-2006

Mobile: (952) = 297-2575

cramer . pollux.chem.umn.edu<= /FONT>

http://pollux.chem.umn.edu/~cr= amer

(website = includes information about the textbook "Essentials

=A0 =A0 of Computational = Chemistry:=A0 Theories and = Models, 2nd Edition")


=

= --Apple-Mail-2--986972513-- From owner-chemistry@ccl.net Tue Feb 21 17:05:00 2006 From: "Sengen Sun sengensun()yahoo.com" To: CCL Subject: CCL: 6-31G basis set Message-Id: <-30967-060221163914-4993-0nYNsKKITqbIeMhN4jwGhQ:-:server.ccl.net> X-Original-From: Sengen Sun Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 Date: Tue, 21 Feb 2006 13:39:05 -0800 (PST) MIME-Version: 1.0 Sent to CCL by: Sengen Sun [sengensun+*+yahoo.com] I would think a molecule of 10 residue peptides is too big to use anything other than HF/3-21G. You may try some pure DFT methods (I have very limited expereience and cannot tell anything more). But it might take forever to finish your calculation with B3LYP. B3LYP is much more time efficient than MP2, especially when MP2 requires a very large basis set. MP2 works significantly better as the basis set gets larger, but you often cannot afford the cost of your time. Good luck. Sengen > From: Goedele Roos > Subject: CCL: 6-31G basis set > Date: Tue, 21 Feb 2006 09:36:11 +0100 (CET) > > -------------------------------------------------------------------------------- > > Sent to CCL by: Goedele Roos [groos/./vub.ac.be] > I have a large system (a 10-residue peptide) and I > was recommended to use rather > HF/3-21G than HF/6-31G, because -so was said to me- > for geometry optimiations, > hf/6-31G works worser than HF/3-21G, unless there > were put some polarization > functions on it... > Is this coirrect, and why? > Another point: there is no point in using larger > basis set with b3LYP. But which > other functional would you recommand than, with > which > larger basis sets are > usefull? > Best regards, goedele > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam > protection around > http://mail.yahoo.com > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From owner-chemistry@ccl.net Tue Feb 21 18:51:00 2006 From: "Cherolyn Brooks cbrooks(_)ucsd.edu" To: CCL Subject: CCL: Summer Workshop - Molecular Simulation and Structure Prediction Message-Id: <-30968-060221173137-2424-613qGKsOPIhxPfFcGPkMag##server.ccl.net> X-Original-From: "Cherolyn Brooks" Date: Tue, 21 Feb 2006 17:31:33 -0500 Sent to CCL by: "Cherolyn Brooks" [cbrooks[-]ucsd.edu] The Center for Theoretical Biological Physics, CTBP, http://ctbp.ucsd.edu, in conjunction with the Center for the Development of Multiscale Modeling Tools for Structural Biology, MMTSB, http:// mmtsb.scripps.edu, will be hosting "Molecular Simulation and Structure Prediction using CHARMM and the MMTSB Tool Set" to be held in La Jolla, CA July 31- August 4, 2006. CTBP has funding available from the National Science Foundation to provide UCSD campus housing and meals for all the invited workshop participants. In addition, funds are available to defray other travel expenses. On-line applications can be submitted at http://ctbp.ucsd.edu/workshops/ workshop_reg.php?id=19 The application deadline is Friday, March 17, 2006. This workshop will provide participants with a solid introduction to the theory and methodology behind current methods of molecular simulations in biophysics: force field development, simulation algorithms, analysis techniques and advanced sampling and simulation methods, and structure prediction/refinement. Participants will be involved in hands-on applications of these principles through tutorials and individual pursuits, conducted with the mentorship of leading researchers in the field of molecular biophysics theory and simulation. The workshop will focus on the implementation of these approaches through the CHARMM macromolecular simulation program and the MMTSB Tool Set. The workshop will be aimed at students, postdocs and faculty who have a background in quantitative areas of biology as well as physics and chemistry. The workshop will be presented by Charles L. Brooks, III, The Scripps Research Institute; Michael Feig, Michigan State University; Wonpil Im, The University of Kansas; Alex MacKerrel, University of Maryland; Lennart Nilsson, Karolinska Institute; and Benoit Roux, The University of Chicago. From owner-chemistry@ccl.net Tue Feb 21 20:50:00 2006 From: "Sue Lam chsue2004+*+yahoo.com" To: CCL Subject: CCL: Question on TDDFT calculation in Gaussian Message-Id: <-30969-060221124234-26493-Sw5dK4FzV/WR+ShAlIIT1A*server.ccl.net> X-Original-From: Sue Lam Content-Transfer-Encoding: 8bit Content-Type: multipart/alternative; boundary="0-1204734510-1140540150=:98972" Date: Tue, 21 Feb 2006 08:42:30 -0800 (PST) MIME-Version: 1.0 Sent to CCL by: Sue Lam [chsue2004^_^yahoo.com] --0-1204734510-1140540150=:98972 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I performed a TDDFT calculation for a triplet transition. Does anyone know how to calculate the % conribution of each excitation in a transition from TDDFT calcuation? Is the sum of the squares of the coefficients for all the excitations in each transition equal to 1? Thanks, Sue --------------------------------- Yahoo! Autos. Looking for a sweet ride? Get pricing, reviews, & more on new and used cars. --0-1204734510-1140540150=:98972 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I performed a TDDFT calculation for a triplet transition.
Does anyone know how to calculate the % conribution of each excitation  in a transition from TDDFT calcuation? Is the sum of the squares of the coefficients for all the excitations in each transition equal to 1?

Thanks,
Sue

Yahoo! Autos. Looking for a sweet ride? Get pricing, reviews, & more on new and used cars. --0-1204734510-1140540150=:98972--