From chemistry-request@server.ccl.net Sun Dec 1 20:02:27 2002 Received: from smtp02.sohu.com ([61.135.132.105]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB212PO05016 for ; Sun, 1 Dec 2002 20:02:26 -0500 Received: from PC12 (unknown [202.127.19.236]) by smtp02.sohu.com (Postfix) with ESMTP id 016846A950 for ; Mon, 2 Dec 2002 09:02:23 +0800 (CST) Date: Mon, 2 Dec 2002 9:2:13 +0800 From: "Liu zhiguo" Reply-To: zgliu@sohu.com To: "CHEMISTRY@ccl.net" Subject: Gromacs vs Amber X-mailer: Foxmail 4.1 [cn] Mime-Version: 1.0 Content-Type: text/plain; charset="GB2312" Content-Transfer-Encoding: 7bit Message-Id: <20021202010223.016846A950@smtp02.sohu.com> Dear CCLers, I'm going to do some work relating to MD.As far as I know,The most commonly used softwares of MD is Gromacs and Amber.I wonder which one is better for me(My research object is nuclear receptor).Could someone make a comparision between them? Best regards! _______________________________________ | | | Liu zhi-guo,Ph.D candidate, | | Shanghai Institute of Materia Medica`| | Phone:021-64311833-615 | | Email:zgliu@sohu.com | |_______________________________________| From chemistry-request@server.ccl.net Mon Dec 2 03:53:53 2002 Received: from mail.ic.sunysb.edu ([129.49.1.4]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB28rrO12305 for ; Mon, 2 Dec 2002 03:53:53 -0500 Received: from postal.ic.sunysb.edu (mail [129.49.1.4]) by mail.ic.sunysb.edu (8.11.6/8.11.6) with SMTP id gB28rmW19447 for ; Mon, 2 Dec 2002 03:53:48 -0500 (EST) Received: from smtp.ic.sunysb.edu ([129.49.1.24]) by postal.ic.sunysb.edu (NAVGW 2.5.2.9) with SMTP id M2002120203534809055 for ; Mon, 02 Dec 2002 03:53:48 -0500 Received: from sparky.ic.sunysb.edu (daemon@sparky.ic.sunysb.edu [129.49.1.3]) by smtp.ic.sunysb.edu (8.11.6/8.11.6) with ESMTP id gB28rmv19444 for ; Mon, 2 Dec 2002 03:53:48 -0500 (EST) Received: from localhost (yonyang@localhost) by sparky.ic.sunysb.edu (8.11.6/8.11.6) with ESMTP id gB28rmq20348 for ; Mon, 2 Dec 2002 03:53:48 -0500 (EST) Date: Mon, 2 Dec 2002 03:53:48 -0500 (EST) From: Yong Liang Yang To: chemistry@ccl.net Subject: Convergence failure in geometry optimization of big molecule Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII hello,dear CCL mates: I am still trying to do the geometry optimization stuff for a big molecule containing boron and nitrogen. The solvation model I was chosing is PCM and I have set the radius of the negative charge center nitrogen using NSFE option with Dump. To increase the tesserae number, TSNUM option is used. After four days CPU time, the job is terminated due to convergence failure, (Convergence criterion not met). And the hint given is: Hint: increase the number of tesserae or change the molecule orientation. But actually I have inceased the tesserae number and I am not sure what is the deal to change the molecule orientation. Does anyone who have some suggestions on the SCF convergency of a big system in solvation model? Thank you all in advances. Yongliang Yang regards From chemistry-request@server.ccl.net Mon Dec 2 07:33:06 2002 Received: from fyserv1.fy.chalmers.se ([129.16.110.66]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2CX5O29727 for ; Mon, 2 Dec 2002 07:33:06 -0500 Received: from fy.chalmers.se (f2-pc31.fy.chalmers.se [129.16.110.231]) by fyserv1.fy.chalmers.se (8.8.8/8.8.8) with ESMTP id NAA10862 for ; Mon, 2 Dec 2002 13:32:11 +0100 (MET) Message-ID: <3DEB5313.401@fy.chalmers.se> Date: Mon, 02 Dec 2002 13:33:23 +0100 From: Patrik Johansson User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: en-us, en MIME-Version: 1.0 To: chemistry@ccl.net Subject: efrag in GAMESS Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Dear all, has anyone out there experience with the EFP modules in GAMESS? I'd really like an example-input made from a runtyp=makefp to get the syntax right. I also wonder if there's a way to make a fragment with less than 3 atoms - and if so - how does one fill in the xyz coordinates in the $efrag? sincerely /Patrik -- Dr. Patrik Johansson, Materials Physics Group, Chalmers, Sweden. From chemistry-request@server.ccl.net Mon Dec 2 10:07:29 2002 Received: from typhoeus.chem.york.ac.uk ([144.32.72.225]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2F7TO00953 for ; Mon, 2 Dec 2002 10:07:29 -0500 Received: from ysbl.york.ac.uk (rasher [144.32.20.17]) by typhoeus.chem.york.ac.uk (8.11.6+Sun/8.11.6) with ESMTP id gB2F7P110050 for ; Mon, 2 Dec 2002 15:07:25 GMT Sender: greaves@ysbl.york.ac.uk Message-ID: <3DEB7634.8BEA3D86@ysbl.york.ac.uk> Date: Mon, 02 Dec 2002 15:03:16 +0000 From: Richard Greaves Reply-To: greaves@ysbl.york.ac.uk Organization: York Structural Biology Laboratory X-Mailer: Mozilla 4.7C-SGI [en] (X11; I; IRIX 6.5 IP32) X-Accept-Language: en MIME-Version: 1.0 To: chemistry@ccl.net Subject: Compiling GAMESS on a SUN under forte v7 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear CCL, I have attached a log file that I generated whilst attempting to compile GAMESS (May 1997 version) using forte v7 fortran compiler and the CC C compiler on a SUN. The code appears to be incorrect. Do you know if later versions of GAMESS have remedied this problem? Is their an email address at Ames that I could contact about this issue? Richard Greaves. From chemistry-request@server.ccl.net Mon Dec 2 11:57:19 2002 Received: from smtp.relc.com ([194.183.162.130]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2GvGO03956 for ; Mon, 2 Dec 2002 11:57:17 -0500 Received: from gw.imbg.org.ua (gw.imbg.org.ua [194.183.176.94]) by smtp.relc.com (8.12.6/8.12.6) with ESMTP id gB2GuhJQ004507 for ; Mon, 2 Dec 2002 18:57:02 +0200 (EET) Received: from violet (violet.imbg.org.ua [10.0.5.80]) by gw.imbg.org.ua (8.12.6/8.11.6) with ESMTP id gB2HBiXX007284 for ; Mon, 2 Dec 2002 19:11:49 +0200 (EET) (envelope-from dikov@imbg.org.ua) From: Dmitry Kovalsky Reply-To: dikov@imbg.org.ua Organization: IMBG To: CHEMISTRY@ccl.net Subject: Cartesian vs DLC Date: Mon, 2 Dec 2002 18:56:20 +0200 User-Agent: KMail/1.4.7 MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1251" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200212021856.20773.dikov@imbg.org.ua> X-AVP-RU: Passed Dear CCL, While performing structure optimization of the enzyme's active site (two identical chains each of 4 residues interconnected by 4 h-bonds) with bound Na+ ion to it, I have noted that the choice of coordinates (CART or DLC) to be used dramatically infulences the final geometry. In case of DLC coordinates the distance between these two chains increses in few Angstroms while Na+ almost immobile. In contrast when the Cartesians is used the Na+ moves a lot and the active site moves very little. Is it Ok? What result should I believe? Sincerely yours, Ph.D. Student Dmitry Kovalsky Institute of Molecular Biology & Genetics 150 Akad. Zabolotnogo Street, Kiev-143, 03143 UKRAINE E-mail: dikov@imbg.org.ua Fax: +380 (44) 266-0759 Tel.: +380 (44) 266-5589 From chemistry-request@server.ccl.net Mon Dec 2 01:50:31 2002 Received: from kdmail2.netcologne.de ([194.8.194.104]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB26oUO09643 for ; Mon, 2 Dec 2002 01:50:31 -0500 Received: from cosmologic.de (xdsl-213-168-119-143.netcologne.de [213.168.119.143]) by kdmail2.netcologne.de (Mirapoint Messaging Server MOS 2.9.3.5) with ESMTP id ABO26738; Mon, 2 Dec 2002 07:50:04 +0100 (CET) Message-ID: <3DEB028E.E411E6E7@cosmologic.de> Date: Mon, 02 Dec 2002 07:49:50 +0100 From: "Dr. Andreas Klamt" Organization: COSMOlogic GmbH&CoKG X-Mailer: Mozilla 4.7 [de] (WinNT; I) X-Accept-Language: en,pdf MIME-Version: 1.0 To: Parthiban , "chemistry@ccl.net" Subject: Re: CCL:Quantum Chemistry in Drug Design References: <20021130014134.M78797@jubilantbiosys.com> Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit My view of this topic is as follows: I am afraid, that realistic modeling of drug receptor interaction is still too demanding for quantum chemical methods, because usually the enzyme is too large, and in addition a lot of conformations and many compounds would have to be sampled. The situation is different if you consider the ADME part: Here we can do good calculations of solubility and many kinds of physiological partitioning properties quite efficiently using DFT methods. The advantage compared with force field models is that you can get much better insight into the real physics using quantum chemistry. For examples see: http://www.cosmologic.de/water_solubility.html or "COSMO-RS: a novel view to physiological solvation and partition questions", Andreas Klamt, Frank Eckert and Martin Hornig, Journal of Computer-Aided Molecular Design 15, 355-365 (2001) and "Prediction of aqueous solubility of drugs and pesticides with COSMO-RS", Andreas Klamt, Frank Eckert, Martin Hornig, Michael E. Beck and Thorsten Bürger, J. Comp. Chem. 23, 275-281 (2002) Andreas Parthiban schrieb: > > Dear Friends: > While several QSAR related techniques and methodologies are appearing in drug > design Journals, very few talk about the more accurate quantum chemical > methods in drug design arena. > > * What are the bottlenecks for the quantum chemical methods to get into the > area of drug design. > > * For small molecules QC methods plays greater role, but for handling drug- > like molecules and handling several thousands of compounds, QC methods do not > see the limelight (correct me if i am wrong). Is the CPU-intensiveness alone > is the reason. Or is there any some conceptual gap in this. [ I hear someone > saying CPU-intensive is the reason and one has to wait for months to get > results ] > > * Based on your experience/insight, can you think of some timeframe, say 5 > years, 10 years down the line, Quantum chemical methods would play a major > role in the area of lead identification/optimization, or would you > say "prediction of future is difficult!". > > I look forward to reading your views. Thanks. > > S. Parthiban > Jubilant Biosys Ltd. > http://www.jubilantbiosys.com > > -= This is automatically added to each message by mailing script =- > CHEMISTRY@ccl.net -- To Everybody | CHEMISTRY-REQUEST@ccl.net -- To Admins > Ftp: ftp.ccl.net | WWW: http://www.ccl.net/chemistry/ | Jan: jkl@ccl.net -- -------------------------------------------------------------------------------- Dr. Andreas Klamt COSMOlogic GmbH&CoKG Burscheider Str. 515 51381 Leverkusen, Germany Tel.: 49-2171-73168-1 Fax: ...-9 e-mail: andreas.klamt@cosmologic.de web: www.cosmologic.de -------------------------------------------------------------------------------- COSMOlogic Your Competent Partner for Computational Chemistry and Fluid Thermodynamics -------------------------------------------------------------------------------- From chemistry-request@server.ccl.net Mon Dec 2 11:20:19 2002 Received: from watusi.nmrfam.wisc.edu ([144.92.227.140]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2GKIO02817 for ; Mon, 2 Dec 2002 11:20:18 -0500 Received: from brown.nmrfam.wisc.edu (brown.nmrfam.wisc.edu [144.92.227.163]) by watusi.nmrfam.wisc.edu (SGI-8.9.3/8.9.3) with ESMTP id KAA49130 for ; Mon, 2 Dec 2002 10:20:08 -0600 (CST) Message-Id: <4.3.2.7.2.20021202100634.00c64bc0@mail.nmrfam.wisc.edu> X-Sender: milo@mail.nmrfam.wisc.edu X-Mailer: QUALCOMM Windows Eudora Version 4.3.2 Date: Mon, 02 Dec 2002 10:20:12 -0600 To: chemistry@ccl.net From: "W. M. Westler" Subject: prop=EPR for ghost atoms Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed I am calculating the hyperfine tensor by using the prop=EPR keyword in Gaussian 98 (A.9). I want to add ghost atoms (Bq) to the molecule and calculate the hyperfine tensor at those sites. The output of EPR only gives values for the atoms in the molecule and not the ghost atoms. In Gaussian 94, using prop=EFG IOp(6/17=2,6/26=4), which is equivalent to EPR, gives the hyperfine tensor for both atoms and ghost atoms. Is there any way to get the hyperfine tensor for ghost atoms in G98 (prop=EFG IOp(6/17=2,6/26=4) gives the same results as prop=EPR )? -- Milo =============================================================================== National Magnetic Resonance Facility at Madison, An NIH-Supported Resource Center W. Milo Westler, Ph.D. EMAIL: milo@nmrfam.wisc.edu NMRFAM Director PHONE: (608)-263-9599 FAX: (608)-263-1722 Senior Scientist and Adjunct Professor 433 Babcock Drive Department of Biochemistry University of Wisconsin-Madison Madison, WI USA 53706-1544 =============================================================================== From chemistry-request@server.ccl.net Mon Dec 2 12:45:27 2002 Received: from magnum.cooper.edu ([199.98.16.4]) by server.ccl.net (8.11.6/8.11.0) with SMTP id gB2HjQO05408 for ; Mon, 2 Dec 2002 12:45:26 -0500 Received: by magnum.cooper.edu id AA29705 (5.65c/IDA-1.4.4 for CHEMISTRY@ccl.net); Mon, 2 Dec 2002 12:47:20 -0500 Date: Mon, 2 Dec 2002 12:47:20 -0500 (EST) From: "Robert Q. Topper" To: CHEMISTRY@ccl.net Cc: TOPPER ROBERT Subject: ECCC9: Abstracts Due Friday December 13, 2002 Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII ECCC9: Abstracts Due Friday December 13, 2002 Dear colleagues, We are pleased to announce that abstracts are being accepted to the 9th Electronic Computational Chemistry Conference. All abstracts must be submitted on or before Friday, December 13. ECCC9 will be held March 1 - March 31, 2003 entirely on the Internet, at http://eccc9.cooper.edu. Abstracts are now being accepted for online submission at the conference website. As in previous years, ECCC9 is a multidisciplinary international conference and covers all aspects of computational and theoretical chemistry, chemical physics and materials science, as well as computational molecular biology, computational and theoretical molecular and atomic physics, visualization, cheminformatics, the history of computational chemistry, and related fields. Participants will be able to view the presentations and discuss them entirely through a web browser. As with the previous eight ECCCs, ECCC9 has NO registration fee and is a completely virtual, online conference Abstracts for all contributions to ECCC9 will be reviewed by the Scientific Organizing Committee (SOC) to insure novelty, scientific value, and appropriateness for inclusion in the conference. All ECCC9 presentations must be in HTML format, with their layout and presentation at the discretion of the authors. Important dates for ECCC9 are: December 13, 2002 - Abstracts of presentations due February 24, 2003 - Final presentations due online March 1, 2003 - Conference begins March 17, 2003 - Interactive Session begins March 24, 2003 - Interactive Session ends April 1, 2003 - ECCC9 ends May 1, 2003 - Submissions to ECCC9 - IJMS Proceedings due The Proceedings of ECCC9 will be published in a special issue of the International Journal of Molecular Sciences (http://www.mdpi.net/ijms/). IJMS is a peer-refereed scientific journal, published online monthly (since September 2001) and CD-ROM yearly. It is indexed and abstracted by several leading indexing and abstracting services, including Chemical Abstracts and CAPLUS, and is physically distributed to various libraries around the world. IJMS articles are freely available to all on the Internet. Please see the ECCC9 RAnnouncementsS page for more information. We look forward to your participation in ECCC9. Robert Q. Topper, Robert Hopkins and Christopher Lent Albert Nerken School of Engineering The Cooper Union for the Advancement of Science and Art 51 Astor Place New York, NY 10003 topper@cooper.edu http://eccc9.cooper.edu From chemistry-request@server.ccl.net Mon Dec 2 13:32:57 2002 Received: from soul.helsinki.fi ([128.214.3.1]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2IWvO06612 for ; Mon, 2 Dec 2002 13:32:57 -0500 Received: from localhost (mpjohans@localhost) by soul.helsinki.fi (8.9.3/8.9.3) with ESMTP id UAA20271; Mon, 2 Dec 2002 20:32:37 +0200 (EET) X-Authentication-Warning: soul.helsinki.fi: mpjohans owned process doing -bs Date: Mon, 2 Dec 2002 20:32:37 +0200 (EET) From: Mikael Johansson X-Sender: To: Dmitry Kovalsky cc: Subject: Re: CCL:Cartesian vs DLC In-Reply-To: <200212021856.20773.dikov@imbg.org.ua> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hello Dimitry and All! On Mon, 2 Dec 2002, Dmitry Kovalsky wrote: > Na+ ion to it, I have noted that the choice of coordinates (CART or DLC) to > be used dramatically infulences the final geometry. > Is it Ok? > What result should I believe? Usually the one with the lower energy is the better geometry. But when modelling an enzyme it might not be what you really want. You could try to do a Cartesian based optimization on the structure optimized with internal coordinates, and vice versa, and see what happens to the structure. As a side note, it's a good idea to specify what program you are using, it usually makes answering easier :-) Have a nice day, Mikael J. http://www.helsinki.fi/~mpjohans/ From chemistry-request@server.ccl.net Mon Dec 2 13:55:09 2002 Received: from simon.cs.cornell.edu ([128.84.154.10]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2It8O07304 for ; Mon, 2 Dec 2002 13:55:08 -0500 Received: from sundial.cs.cornell.edu (sundial.cs.cornell.edu [128.84.96.115]) by simon.cs.cornell.edu (8.11.3/8.11.3/R-3.7) with ESMTP id gB2It7D22400 for ; Mon, 2 Dec 2002 13:55:07 -0500 (EST) Received: from ringding.cs.cornell.edu (ringding.cs.cornell.edu [128.84.96.109]) by sundial.cs.cornell.edu (8.11.3/8.11.3/M-3.10) with ESMTP id gB2It6924174 for ; Mon, 2 Dec 2002 13:55:06 -0500 (EST) Received: from localhost (alfredo@localhost) by ringding.cs.cornell.edu (8.11.3/8.11.3/C-3.2) with SMTP id gB2It5n15974 for ; Mon, 2 Dec 2002 13:55:06 -0500 (EST) Date: Mon, 2 Dec 2002 13:55:05 -0500 (EST) From: Alfredo Cardenas To: chemistry@ccl.net Subject: steepest descent path program Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hello all, Does anybody have a code to compute the steepest descent path between two minima states (coordinates)? Or could you point me to anyone that could have such a code? Thanks a lot, Alfredo Cardenas Dept. Computer Science Cornell University alfredo@cs.cornell.edu From chemistry-request@server.ccl.net Mon Dec 2 20:09:06 2002 Received: from viper.abbott.com ([130.36.44.77]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB3196O13743 for ; Mon, 2 Dec 2002 20:09:06 -0500 Received: from abtmr1.abbott.com (abtmr1.abbott.com [10.254.245.1]) by viper.abbott.com (Switch-2.1.4/Switch-2.1.0) with ESMTP id gB31A9u28857 for ; Mon, 2 Dec 2002 19:10:14 -0600 (CST) Received: from abtapn561.northamerica.intra.abbott.com (localhost [127.0.0.1]) by abtmr1.abbott.com (Switch-2.0.6/Switch-2.0.6) with ESMTP id gB3191j07562 for ; Mon, 2 Dec 2002 19:09:01 -0600 (CST) To: chemistry@ccl.net Cc: james.metz@abbott.com Subject: CCL:constrained multiple linear regression X-Mailer: Lotus Notes Release 5.0.5 September 22, 2000 Message-ID: From: james.metz@abbott.com Date: Mon, 2 Dec 2002 19:08:14 -0600 X-MIMETrack: Serialize by Router on ABTAPN561/ESVR/ABBOTT(Release 5.0.5 |September 22, 2000) at 12/02/2002 07:04:51 PM, Serialize complete at 12/02/2002 07:04:51 PM MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="=_alternative 0006511186256C84_=" This is a multipart message in MIME format. --=_alternative 0006511186256C84_= Content-Type: text/plain; charset="us-ascii" CCL, I am looking for software (preferably free or inexpensive, code is OK) that will allow me to perform constrained multiple linear regression. For example, I have an equation of the form: Y = a*X1 + b*X2 + c*X3 + d where: Y is the dependent variable X1 - X3 are independent variables (more than 1) a, b, c are coefficients d is a constant I must be able to constrain the range of values for a, b, c, and d during the fitting. Please note that programs such as JMP, KalediaGraph, GraphPad Prism, and several other packages do NOT allow the user to do this, especially for more than one independent variable. Can someone suggest a program or code to do this? Thank you, Jim Metz James T. Metz, Ph.D. Research Investigator Chemist GPRD R46Y AP10-2 Abbott Laboratories 100 Abbott Park Road Abbott Park, IL 60064-6100 U.S.A. Office (847) 936 - 0441 FAX (847) 935 - 0548 james.metz@abbott.com --=_alternative 0006511186256C84_= Content-Type: text/html; charset="us-ascii"
CCL,

        I am looking for software (preferably free or inexpensive, code is OK) that will allow me to perform
constrained multiple linear regression.  For example, I have an equation of the form:

        Y = a*X1 + b*X2 + c*X3 + d

        where: Y is the dependent variable
                  X1 - X3 are independent variables (more than 1)
                a, b, c are coefficients
                        d is a constant

        I must be able to constrain the range of values for a, b, c, and d during the fitting.

        Please note that programs such as JMP, KalediaGraph, GraphPad Prism, and several other packages do
NOT allow the user to do this, especially for more than one independent variable.

        Can someone suggest a program or code to do this?

        Thank you,
        Jim Metz


James T. Metz, Ph.D.
Research Investigator Chemist

GPRD R46Y AP10-2
Abbott Laboratories
100 Abbott Park Road
Abbott Park, IL  60064-6100
U.S.A.

Office (847) 936 - 0441
FAX    (847) 935 - 0548

james.metz@abbott.com
 --=_alternative 0006511186256C84_=-- From chemistry-request@server.ccl.net Mon Dec 2 13:15:01 2002 Received: from sandmail01.sd.accelrys.com ([209.120.169.30]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2IF1O06114 for ; Mon, 2 Dec 2002 13:15:01 -0500 To: chemistry@ccl.net Cc: elizabethc@accelrys.com Subject: Accelrys Training Workshops for February MIME-Version: 1.0 X-Mailer: Lotus Notes Release 5.0.9a January 7, 2002 Message-ID: From: jnauss@accelrys.com Date: Mon, 2 Dec 2002 10:14:58 -0800 X-MIMETrack: Serialize by Router on sandmail01/Server/Accelrys(Release 5.0.9 |November 16, 2001) at 12/02/2002 10:15:00 AM, Serialize complete at 12/02/2002 10:15:00 AM Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id gB2IF1O06115 Accelrys Inc. are holding the following training workshops at locations in the US and Europe during February 2003. These events are designed to help you get more value from your Accelrys software, helping you to better accomplish your research goals. Costs are $600 per day for commercial, $450 per day for government and $330 per day for academic. Cheminformatics: DS Accord Enterprise OMNIS Client 3rd February in Cambridge, UK RS3 for Excel 4th February in Cambridge, UK RS3 Inventory 5th February in Cambridge, UK OMMM 6th February in Cambridge, UK DIVA 7th February in Cambridge, UK DS Accord Biostore Database Administration 10th-11th February in Cambridge, UK For course details and registration see: http://www.accelrys.com/training/cheminf/schedule.html Macromolecular Modeling: Biomolecular NMR: Processing and Analysis 11th-12th February in San Diego, CA Biomolecular NMR: Structure Determination 13th ?14th February in San Diego, CA For course details and registration see: http://www.accelrys.com/training/macro/schedule.html Bioinformatics: Wisconsin Package and SeqLab 17th-18th February in Cambridge, UK SeqWeb 19th February in Cambridge, UK DS Gene Workshop 20th-21st February in Cambridge, UK For course details and registration see: http://www.accelrys.com/training/bioinf/schedule.html -- Jeffrey L. Nauss, Ph.D. Manager, Discovery Studio Training Accelrys Inc. 9685 Scranton Road San Diego, CA 92121-3752 Phone: 858-799-5555 Fax: 858-799-5100 E-mail: jnauss@accelrys.com http://www.accelrys.com/training From chemistry-request@server.ccl.net Mon Dec 2 13:10:55 2002 Received: from smtp.uh.cu ([216.72.24.105]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB2IArO06001 for ; Mon, 2 Dec 2002 13:10:54 -0500 Received: from fq.uh.cu (canaria.qct.fq.oc.uh.cu [192.168.3.4]) by smtp.uh.cu (Postfix) with ESMTP id 268833401E for ; Mon, 2 Dec 2002 14:10:29 -0500 (CST) Received: from Yigni (espectro.qct.fq.oc.uh.cu [192.168.3.44]) by fq.uh.cu (8.9.0/8.8.7) with SMTP id NAA07787 for ; Mon, 2 Dec 2002 13:12:06 -0500 Message-Id: <200212021812.NAA07787@fq.uh.cu> From: "=?ISO-8859-1?Q?Reynier_Suardiaz_del_R=EDo?=" Organization: =?ISO-8859-1?Q?Qu=EDmica_-_F=EDsica?= To: CHEMISTRY@ccl.net Date: Sat, 2 Dec 2000 13:10:05 +0000 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: experimental rotational barrier Priority: normal In-reply-to: <200212021856.20773.dikov@imbg.org.ua> X-mailer: Pegasus Mail for Windows (v3.01d) I'm trying to build the parameters for a MD simulation of an epoxidic sistem but a have not found any experimental data of the raotational barrier for the dihedral angle H1-C2-C3-C4 in the following sistem: _ I'll be gratefull of any suggestion. _____________________________________________ Reynier Suardiaz del Rio Department of Physical Chemistry Chemistry Faculty University of Havana email: reynier@fq.uh.cu, reynier_s@yahoo.com _____________________________________________ From chemistry-request@server.ccl.net Mon Dec 2 22:20:57 2002 Received: from ccl.net ([192.148.249.4]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id gB33KvO16130 for ; Mon, 2 Dec 2002 22:20:57 -0500 Received: from arlen.ccl.net (arlen.ccl.net [192.148.249.10]) by ccl.net (8.11.6+Sun/8.11.6/OSC 2.1) with ESMTP id gB33Kvb21034; Mon, 2 Dec 2002 22:20:57 -0500 (EST) Date: Mon, 2 Dec 2002 22:20:57 -0500 (EST) From: Jan Labanowski To: james.metz@abbott.com cc: Jan Labanowski , chemistry@ccl.net Subject: Re: CCL:constrained multiple linear regression In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII You need a program which does regression through origin (i.e., no intercept regression. Check references (sorry, it was ages ago when I looked at it): K.W. Smillie - "An introduction to regression and correlation" Academic Press, NY 1966 S.R. Searle - "Linear models", J.Wiley & Sons, NY 1971 I once write my own routine to do it, but it is in a terribly ancient Fortran (yes... remember VAXes? in the early 80-ties?). You have a regression Y = X*b + e When freeft = 1, then we fit an intercept and the X contains a column of 1's When freeft = 0, then we have regression throught origin and X does not have a column of 1's (Note, you can always subtract a constant from all values of Y if you want to fix intercept at some value. Likewise, if you want to fix some b, you just subtract the b(j)*X(i,j) from Y(i) ) Y - column vector of "experimental values" X - matrix storing values of independent variables: X(i,j) - value j-th independent variable for i-th experimental b - column vector of regression coefficients b(i) - regresion coefficients corresponding to i-th independent variable e - column vector of residuals Nexp - number of experimental points Nparam - number of fitted regression coefficients If you have statistical weights, W (i.e., reciprocals of variance for each experimental point), calculate their mean Wmean = (sum of Wi)/Nexp and calculate normalized roots of weights as: Wroot(i) = sqrt(W(i)/Wmean) Then rescale Y and X by Wroots as: Y(i) = Y(i) * Wroot(i) (for i = 1, Nexp) X(i,j) = X(i,j) * Wroot(i) (for i = 1, Nexp and for j = 1, Nparam) Then calculate: C XTX - on output inverse of X'X matrix (inside the routine, C before inversion, it stores elements of X'X) C XTY - X'Y matrix C YTY - Y'Y - sum of squares of Yexp(i) C rXiXj - array of correlation coefficients between independent C variables Xi and Xj C rYXi - array of correlation coefficients between dependent var- C iable Y and independent variable Xi C SSE - sum of squares of residuals C SSE = sum of (Yexp(i) - Ytheor(i))**2 (i.e Y'Y - b'X'Y) C both for freeft = 0 & freeft = 1 C C SSR - sum of sqyares about regression C SSR = sum of (Ytheor(i) - Ymean)**2 C or b'X'Y - Nexp*Ymean*Ymean when freeft = 1 C SSR = b'X'Y when freeft = 0 C C SST - total sum of squares C SST = (Yexp(i) - Ymean)**2 C or Y'Y - Nexp*Ymean*Ymean when freeft = 1 C SST = Y'Y when freeft = 0 C C MSE - mean square of errors { MSE = SSE/(Nexp-Nparam) } C C MSR - mean square of regression C MSR = SSR/Nparam for freeft = 0 C MSR = SSR/(Nparam - 1) for freeft = 1 C C b - regression coefficients b = inv(X'X)*X'Y C inv(X'X) is the inverse of X'X C C seb - standard errors of regression coefficients C seb[i] = SQRT(MSE/inv(X'X[i,i])) C where inv(X'X[i,i]) is the i,i-element of the C inverse of X'X matrix C C seYth - standard errors of regression values C seYth[i] = SQRT(MSE*Xi'*inv(X'X)*Xi) where C Xi is the i-th column of X matrix and C inv(X'X) is the inverse of X'X matrix C C Rmul - multiple regression coefficient that is SQRT(Rmul2) or C in case of Rmul2 < 0 (sometimes in no-intercept model) C Rmul is set to 0 C C Rmul2 - Square of multiple correlation coefficient C Rmul2 = SSR/SST C C Xmean - array of means of independent variables C C Ymean - mean of Yexp (and Ytheor) C C F - F statistics F = MSR/MSE C C t - t statistics for each regression coefficient for testing C hypothesis that bstar(i) = 0 C t(i) = b[i]/seb[i] C C The ugly Fortan follows INTEGER*2 Nexp, Nparam, df, i, j, k, l, errflg, mode,freeft REAL*8 X(MAXEXP,MAXPAR), Yexp(MAXEXP), Ytheor(MAXEXP), 1 W(MAXEXP), Wroots(MAXEXP), Wmean, t(MAXPAR), 2 rXiXj(MAXPAR,MAXPAR), rYXi(MAXPAR), Xmean(MAXPAR), 3 XTX(MAXPAR,MAXPAR), XTY(MAXPAR), SSE, SSR, SST, 4 seYth(MAXEXP), seb(MAXPAR), b(MAXPAR), Rmul2, 5 Rmul, F, MSR, MSE, Ymean, YTY, y, s, s1, s2, aux C Read in the Yexp, X, W and freeft (find the return statements and C change them to STOP) or make a subroutine out of this code C .................... C C C Calculate Wmean for relative weights C Wmean = 0.0D0 DO 10 i = 1, Nexp Wmean = Wmean + W(i) 10 CONTINUE Wmean = Wmean/Nexp C C Rescale Yexp and X for weighted least squares C 15 DO 20 i = 1, Nexp Wroots(i) = SQRT(W(i)/Wmean) 20 CONTINUE DO 40 i = 1, Nexp s = Wroots(i) Yexp(i) = Yexp(i) * s DO 30 j = 1, Nparam X(i,j) = X(i,j) * s 30 CONTINUE 40 CONTINUE C C Calculate Y'Y and Ymean C YTY = 0.0D0 Ymean = 0.0D0 DO 42 i = 1, Nexp y = Yexp(i) YTY = YTY + y*y Ymean = Ymean + y 42 CONTINUE Ymean = Ymean/Nexp C C Calculate Xmean C DO 44 i = 1, Nparam s = 0.0D0 DO 43 j = 1, Nexp s = s + X(j,i) 43 CONTINUE Xmean(i) = s/Nexp 44 CONTINUE C C Calculate X'X C DO 70 i = 1, Nparam DO 60 j = i, Nparam s = 0.0D0 DO 50 k = 1, Nexp s = s + X(k,i)*X(k,j) 50 CONTINUE XTX(i,j) = s IF(i .ne. j) XTX(j,i) = s 60 CONTINUE 70 CONTINUE C C Calculate X'Y C DO 90 i = 1, Nparam s = 0.0D0 DO 80 j = 1, Nexp s = s + X(j,i) * Yexp(j) 80 CONTINUE XTY(i) = s 90 CONTINUE C C Calculate correlation coefficients rXiXj and rYXi C DO 92 i = 1, Nparam s1 = Nexp * Xmean(i)**2 DO 91 j = i, Nparam s2 = Nexp * Xmean(j)**2 s = XTX(i,j) - Nexp * Xmean(i) * Xmean(j) IF (ABS(XTX(i,i) - s1) .LT. 0.00001D0*s1 .OR. 1 ABS(XTX(j,j) - s2) .LT. 0.00001D0*s2) THEN rXiXj(i,j) = 0.0D0 ELSE aux = (XTX(i,i)-s1)*(XTX(j,j)-s2) IF(aux .le. 1.0d-30)THEN errflg = 30 return end if rXiXj(i,j) = s / SQRT((XTX(i,i)-s1)*(XTX(j,j)-s2)) ENDIF IF(i .NE. j) rXiXj(j,i) = rXiXj(i,j) 91 CONTINUE 92 CONTINUE DO 93 i = 1, Nparam s = XTY(i) - Nexp * Xmean(i) * Ymean s1 = Nexp * Xmean(i)**2 s2 = Nexp * Ymean*Ymean IF (ABS(XTX(i,i) - s1) .LT. 0.00001D0*s1 .OR. 1 ABS(YTY - s2) .LT. 0.00001D0*s2) THEN rYXi(i) = 0.0D0 ELSE aux = (XTX(i,i)-s1)*(YTY-s2) if(aux .le. 1.0D-30)THEN errflg = 30 return end if rYXi(i) = s / SQRT((XTX(i,i)-s1)*(YTY-s2)) ENDIF 93 CONTINUE C C Invert X'X C CALL syminv(Nparam, XTX, s, errflg) IF(errflg .ne. 0)GOTO 999 C C Calculate regression coefficients b = inv(X'X)*X'Y C DO 110 i = 1, Nparam s = 0.0D0 DO 100 j = 1, Nparam s = s + XTX(i,j) * XTY(j) 100 CONTINUE b(i) = s 110 CONTINUE C C Calculate Ytheor (values of y given by regression) Ytheor = X*b C DO 130 i = 1, Nexp s = 0.0D0 DO 120 j = 1, Nparam s = s + b(j) * X(i,j) 120 CONTINUE Ytheor(i) = s 130 CONTINUE C C Calculate SST C IF(freeft .eq. 1)THEN SST = YTY - Nexp*Ymean*Ymean ELSE SST = YTY END IF C C Calculate sum of squares of residuals (SSE) C SSE = YTY DO 140 i = 1, Nparam SSE = SSE - XTY(i)*b(i) 140 CONTINUE IF(SSE .lt. -5.0d-10*YTY)THEN errflg = 2 RETURN ELSE IF(SSE .lt. 0.0D0)THEN SSE = 0.0D0 END IF C C Calculate Rmul2, and SSR (sum of squares due C to regression ) C df = Nexp - Nparam SSR = SST - SSE if(abs(SST) .gt. 1.0D-30)then Rmul2 = SSR/SST else Rmul2 = 0.0D0 end if C mean square due to regression (MSR) and mean square of error (MSE) IF(freeft .eq. 0)THEN MSR = SSR/Nparam ELSE MSR = SSR/(Nparam - 1) END IF MSE = SSE/df C Calculate F statistic (F), multiple regression coefficient (Rmul) if(ABS(MSE) .gt. 1.0D-30)then F = MSR/MSE else F = 0.0D0 end if IF(Rmul2 .gt. 0.0D0)THEN Rmul = SQRT(Rmul2) ELSE Rmul = 0.0D0 END IF C Calculate t-statistic and standard error for each coefficient DO 150 i = 1, Nparam IF(XTX(i,i)*MSE .gt. 1.0D-30)THEN s = SQRT(MSE * XTX(i,i)) seb(i) = s t(i) = b(i)/s ELSE seb(i) = 0.0D0 t(i) = 0.0D0 END IF 150 CONTINUE C Calculate standard error for each value of Ytheor DO 180 i = 1, Nexp s1 = 0.0D0 DO 170 j = 1, Nparam s = 0.0D0 DO 160 k = 1, Nparam s = s + XTX(j,k) * X(i,k) 160 CONTINUE s1 = s1 + s * X(i,j) 170 CONTINUE if(MSE*s1 .le. 1.0D-30)then seYth(i) = 0.0D0 else seYth(i) = SQRT(MSE*s1) end if 180 CONTINUE C Restore original values of Yexp and X DO 200 i = 1, Nexp s = 1.0D0/Wroots(i) Yexp(i) = Yexp(i) * s DO 190 j = 1, Nparam X(i,j) = X(i,j) * s 190 CONTINUE 200 CONTINUE On Mon, 2 Dec 2002 james.metz@abbott.com wrote: > Date: Mon, 2 Dec 2002 19:08:14 -0600 > From: james.metz@abbott.com > To: chemistry@ccl.net > Cc: james.metz@abbott.com > Subject: CCL:constrained multiple linear regression > > CCL, > > I am looking for software (preferably free or inexpensive, code is > OK) that will allow me to perform > constrained multiple linear regression. For example, I have an equation > of the form: > > Y = a*X1 + b*X2 + c*X3 + d > > where: Y is the dependent variable > X1 - X3 are independent variables (more than 1) > a, b, c are coefficients > d is a constant > > I must be able to constrain the range of values for a, b, c, and d > during the fitting. > > Please note that programs such as JMP, KalediaGraph, GraphPad > Prism, and several other packages do > NOT allow the user to do this, especially for more than one independent > variable. > > Can someone suggest a program or code to do this? > > Thank you, > Jim Metz > > > James T. Metz, Ph.D. > Research Investigator Chemist > > GPRD R46Y AP10-2 > Abbott Laboratories > 100 Abbott Park Road > Abbott Park, IL 60064-6100 > U.S.A. > > Office (847) 936 - 0441 > FAX (847) 935 - 0548 > > james.metz@abbott.com > Jan K. Labanowski | phone: 614-292-9279, FAX: 614-292-7168 Ohio Supercomputer Center | E-mail: jkl@ccl.net 1224 Kinnear Rd, | http://www.ccl.net/~jkl Columbus, OH 43212-1163 | http://www.ccl.net/ http://asdn.net/