From smori@utsc.s.u-tokyo.ac.jp  Thu Jun 13 02:53:46 1996
Received: from utsc.s.u-tokyo.ac.jp  for smori@utsc.s.u-tokyo.ac.jp
	by www.ccl.net (8.7.5/950822.1) id CAA29246; Thu, 13 Jun 1996 02:45:21 -0400 (EDT)
Received: by utsc.s.u-tokyo.ac.jp (8.7.5+2.6Wbeta6/TISN-1.3/R2)
	id PAA16823; Thu, 13 Jun 1996 15:45:02 +0900 (JST)
Date: Thu, 13 Jun 1996 15:45:02 +0900 (JST)
From: Seiji Mori <smori@utsc.s.u-tokyo.ac.jp>
Message-Id: <199606130645.PAA16823@utsc.s.u-tokyo.ac.jp>
To: chemistry@www.ccl.net
Subject: AIMPAC install problem for AIX



 Dear netters,

Thank you for your efforts of the information of plotting packages
for Laplacian-rho. 
 I am now trying to use gnuplot for Macintosh as 
the plotting program. Do you think that it is easy to convert from 
the output of contor program in AIMPAC to plot 
contour maps by gnuplot without the 
modification of source program?

Main topic is following.

 I am now going to install AIMPAC 94, which Prof.Bader et al gave me in 
the last year in our IBM RS/6000 AIX ver.3.2.5 .
  ext94b,bubble,cubev,  proaimv can be compiled, however,other 
files such as  contor, grdveci was failed to be compiled.
Error messages are following.
---
(EXAMPLE)
root@nec$make grdvec
        xlf -O  grdvec.f -o grdvec
"grdvec.f", line 16.12: 1514-022 (W) Variable iwfn is in or is equivalenced to .
"grdvec.f", line 16.23: 1514-022 (W) Variable ivec is in or is equivalenced to .
"grdvec.f", line 16.34: 1514-022 (W) Variable iout is in or is equivalenced to .
** grdvec   === End of Compilation 1 ===
** euler   === End of Compilation 2 ===
** gaus2   === End of Compilation 3 ===
** gaus4   === End of Compilation 4 ===
** grockle   === End of Compilation 5 ===
** hess   === End of Compilation 6 ===
** makname   === End of Compilation 7 ===
** nuclei   === End of Compilation 8 ===
** number   === End of Compilation 9 ===
"grdvec.f", line 1076.12: 1514-022 (W) Variable modeg is in or is equivalenced .
"grdvec.f", line 1076.20: 1514-022 (W) Variable modes is in or is equivalenced .
** rdwfn   === End of Compilation 10 ===
** shftvc   === End of Compilation 11 ===
** tqlgrm   === End of Compilation 12 ===
** trace   === End of Compilation 13 ===
** tredig   === End of Compilation 14 ===
** trudge   === End of Compilation 15 ===
1501-510  Compilation successful for file grdvec.f.
0706-317 ERROR: Unresolved or undefined symbols detected:
                 Symbols in error (followed by references) are
                 dumped to the load map.
                 The -bloadmap:<filename> option will create a load map.
 .plot
 .plots
make: 1254-004 The error code from the last command is 8.

Make Quitting.
---
( profil)
 .dgemm
---
(contor)
                 The -bloadmap:<filename> option will create a load map.
 .dashdf
 .flush
 .plot
 .plots
make: 1254-004 The error code from the last command is 8.
---
What should I solve this problem?

 Seiji Mori


####################################################
   Seiji Mori
  Graduate student
 Lab. of Physical Organic Chemistry
  Department of Chemistry
 The University of Tokyo
 Hongo 7-3-1, Bunkyou-ku, Tokyo 113,
  JAPAN.
 email:smori@utsc.s.u-tokyo.ac.jp
---
http://www.chem.s.u-tokyo.ac.jp/Students/smori.html
#################################################### 


From b_duke@lacebark.ntu.edu.au  Thu Jun 13 03:12:54 1996
Received: from lacebark.ntu.edu.au  for b_duke@lacebark.ntu.edu.au
	by www.ccl.net (8.7.5/950822.1) id BAA29087; Thu, 13 Jun 1996 01:58:59 -0400 (EDT)
From: <b_duke@lacebark.ntu.edu.au>
Received: by lacebark.ntu.edu.au (AIX 3.2/UCB 5.64/4.03)
          id AA10657; Fri, 14 Jun 1996 15:04:18 +1100
Message-Id: <9606140404.AA10657@lacebark.ntu.edu.au>
Subject: G94: Running G94 from cgi script.
To: CHEMISTRY@www.ccl.net (chemistry)
Date: Fri, 14 Jun 1996 15:04:17 +1100 (EETDT)
X-Mailer: ELM [version 2.4 PL25]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit


Hi all,

I have been setting up web pages that run Gaussian from a perl
cgi script. There are two actions :- (1) the job is run in a
NQS queue and the output either e-mailed to the user or printed;
(2) the job is run interactively and the results sent to the
browser. Both actions (1) and (2) work fine on a server on
a Dec Alpha. Only (1) works on a IBM RS6K. Both servers are
running NCSA httpd. 

The interactive scripts on the RS6K just do nothing and this problem
has been worrying me for many weeks. Here is the problem stripped
down to just bare essentials. 

The html page:-
<HTML> <HEAD>
<TITLE>Test1</TITLE>
</HEAD> <BODY>
<H1 ALIGN=center>Test1.</H1>
<FORM METHOD="post" ACTION="/cgi-bin/test1.pl">
<INPUT TYPE="submit" VALUE="Submit">
<INPUT TYPE="reset" VALUE="Reset"><P>
</FORM>
</BODY> </HTML>

sends no data (as the main page would) and just posts to a cgi
script. If the script is

#!/usr/local/bin/perl
#
$ENV{'g94root'} = '/var';
$ENV{'GAUSS_EXEDIR'} = '/var/g94/bsd:/var/g94';
$ENV{'GMAIN'} = '/var/g94/bsd:/var/g94';
$ENV{'PATH'} = '/usr/bin:/etc:/usr/sbin:/usr/local/bin:/usr/lbin:/usr/bin/X11:/sbin:/var/g94/bsd:/var/g94:.';

system ("testrt \"#N \" > t.log");

print "Content-type: text/html\n\n";
print "<TITLE>Test</TITLE>\n\n";
print "<H1 ALIGN=center>Test done</H1>\n\n";
print "</BODY></HTML>\n";

it just returns "Test done" and leaves t.log created but empty. Note 
this does not even run G94. It runs the small gaussian utility testrt.
This script runs at the prompt fine and puts the testrt output into
t.log.

In order to try to understand this further I tried stuff like this 
script:-

#!/usr/local/bin/perl
#
$ENV{'g94root'} = '/var';
$ENV{'GAUSS_EXEDIR'} = '/var/g94/bsd:/var/g94';
$ENV{'GMAIN'} = '/var/g94/bsd:/var/g94';
$ENV{'PATH'} = '/usr/bin:/etc:/usr/sbin:/usr/local/bin:/usr/lbin:/usr/bin/X11:/sbin:/var/g94/bsd:/var/g94:.';

$job = "t.job";

# write the job file
open(JOB, "> $job") || die "Damn! Can't create $job!\n";
print JOB "#!/bin/ksh\n";
print JOB "whence testrt > t.log\n";
print JOB "testrt \"#N \" >> t.log 2>&1\n";
close(JOB);

chmod 0775, $job;
system "./$job";

print "Content-type: text/html\n\n";
print "<TITLE>Test</TITLE>\n\n";
print "<H1 ALIGN=center>Test done</H1>\n\n";
print "</BODY></HTML>\n";

Here a korn shell script is created and run. The whence line gives
the correct result, while again nothing goes into t.log. The shell
script and the full perl script both run correctly from the prompt.
The script can contain all sorts of commands that work fine. I have used
printenv to check the environment and so on. I have moved other code into
the g94 directory with the same permissions and ownership as the g94
executables. They work. The httpd runs as user web. I had it running
as nobody before. The results are the same. The server on the Dec is running
as web. Remember that there the same scripts run just fine.The cgi script 
from the web leaves this error in the error_log file:-

 ./t.job[3]: 7496 Killed

If I use C-shell, I get no error message, but it does not do anything. 
Clearly the testrt call on line 3 has been started as process 7496, 
but then killed.

Can anyone:-

1. Suggest what is wrong?
2. Suggest how to get more information about the "Killed" message?
   I would prefer not to have to alter the Gaussian code.
3. Run the above on a web server running on a RS6K (AIX 3.2 preferably)?

I would really welcome some feedback on this. It is driving me mad and
we do need to get both batch and interactive use from the web server on 
the RS6K. I want to use the Dec alpha for research when I have finished
marking examinations next week.

Sorry this has been so long. I think it is relevant to CCL as it seems
to be a G94 problem and the use of CC code off web servers has some real
educational value. My students love it. It is just working on the wrong
damn machine!

Cheers, Brian.
-- 
        Associate Professor Brian Salter-Duke (Brian Duke)
School of Mathematical and Physical Sciences, Northern Territory University,
  Darwin, NT 0909, Australia.  Phone 08-89466702. Fax 08-89410460
e-mail: b_duke@lacebark.ntu.edu.au  WWW http://lacebark.ntu.edu.au/chemistry 

From Steven.Creve@chem.kuleuven.ac.be  Thu Jun 13 03:53:48 1996
Received: from hartree.quantchem.kuleuven.ac.b  for Steven.Creve@chem.kuleuven.ac.be
	by www.ccl.net (8.7.5/950822.1) id CAA29322; Thu, 13 Jun 1996 02:59:11 -0400 (EDT)
Received: by hartree.quantchem.kuleuven.ac.be id AA27340
  (5.67b/IDA-1.5 for Computational Chemistry List <chemistry@www.ccl.net>); Thu, 13 Jun 1996 08:59:05 +0100
Date: Thu, 13 Jun 1996 08:59:05 +0100 (NFT)
From: Steven Creve <Steven.Creve@chem.kuleuven.ac.be>
To: Computational Chemistry List <chemistry@www.ccl.net>
Subject: CCL: g94 error
Message-Id: <Pine.A32.3.91.960613085516.26466A-100000@hartree.quantchem.kuleuven.ac.be>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


Hi

Does anyone know what the following error message of G94 means?:

File extend in NtrExt1 failed.
File extend in NtrExt1 failed
File extend in NtrExt1 failed: Invalid argument


I got this when running the following job:

$RunGauss
%mem=6500000
%chk=pf2
# QCISD/gen test scf=direct DENSITY=current

PF2   6-311+G(2df,p) fully uncontracted s-part on P

0 2
P
F     1     r1
F     1     r1        2     a1

   r1          1.6025
   a1         98.5904

@/home/steven/BASIS/g94/pople/p.us6-311+G2dfp/N
****
F 0
6-311+G(2df,p)
****


The p.us6-311+G2dfp user-defined basis set is a 6-311+G(2df,p) on P, 
where the s-functions have been uncontracted. This yields a total number 
of basis functions for PF2 of 116.


--------------------------------------------------------------------------
Steven Creve                       steven.creve@chem.kuleuven.ac.be
Labo Quantumchemie                 steven@hartree.quantchem.kuleuven.ac.be
Celestijnenlaan 200F
3001-HEVERLEE                      tel: (32) (16) 32 73 93
BELGIUM                            fax: (32) (16) 32 79 92


From laaksone@csc.fi  Thu Jun 13 04:53:55 1996
Received: from voxopm.minedu.fi  for laaksone@csc.fi
	by www.ccl.net (8.7.5/950822.1) id DAA29602; Thu, 13 Jun 1996 03:54:27 -0400 (EDT)
Received: (from laaksone@localhost) by voxopm.minedu.fi (8.7.5/8.6.11+CSC-2.0) id KAA18875; Thu, 13 Jun 1996 10:54:26 +0300 (EET DST)
Content-Transfer-Encoding: 8bit
Date: Thu, 13 Jun 1996 10:54:26 +0300 (EET DST)
From: Leif Laaksonen <laaksone@csc.fi>
X-Sender: laaksone@voxopm.minedu.fi
To: chemistry@www.ccl.net
Subject: summary of my Windows NT in chemistry query
Message-ID: <Pine.SGI.3.91.960613104915.18661A-100000@voxopm.minedu.fi>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



Hi,

Some time ago I mailed a query about the interest and role of Windows NT
in the chemistry software field. There was no big discussion this time. I
guess this issue is not too interesting for the general computational
chemistry community. 

If you are interested to read the summary you can find it at:

http://laaksonen.csc.fi/docs/pcunix1.html

Regards,

-leif laaksonen

-------------------------------------------------------------------
Leif Laaksonen                     |  
Center for Scientific Computing    | Phone:      358 0 4572378
P.O. Box 405                       | Mobile:     358 400425203
FIN-02101 Espoo                    | Telefax:    358 0 4572302
FINLAND                            | Mail:  Leif.Laaksonen@csc.fi
---------URL: http://laaksonen.csc.fi/leif.laaksonen.html----------




From owner-chemistry@ccl.net  Thu Jun 13 05:53:49 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id FAA00515; Thu, 13 Jun 1996 05:41:25 -0400 (EDT)
Received: from comsun.rz.uni-regensburg.de  for herbert.homeier@rchs1.chemie.uni-regensburg.de
	by bedrock.ccl.net (8.7.5/950822.1) id FAA27469; Thu, 13 Jun 1996 05:41:22 -0400 (EDT)
Received: from rchs1.chemie.uni-regensburg.de by comsun.rz.uni-regensburg.de with SMTP id AA25167
  (5.65c/IDA-1.4.4 for <chemistry@ccl.net>); Thu, 13 Jun 1996 11:40:28 +0200
Received: by rchs1.chemie.uni-regensburg.de (4.1/URRZ-sub (1.5))
	id AA25941; Thu, 13 Jun 96 11:40:18 +0200
Date: Thu, 13 Jun 96 11:40:18 +0200
From: Herbert Homeier t4720 <herbert.homeier@chemie.uni-regensburg.de>
Message-Id: <9606130940.AA25941@rchs1.chemie.uni-regensburg.de>
To: chemistry@ccl.net, friedric@techfak.uni-bielefeld.de,
        c5008@rchs1.chemie.uni-regensburg.de
Subject: Re: CCL: Smallest rectangular box
Content-Type: text/plain
Mime-Version: 1.0
X-Mailer: NCSA Mosaic 2.6 on Sun
X-Url: http://rchs1.chemie.uni-regensburg.de/%7Emaillist/data/0029.html


> Dear CCL'ers,
> I've got the problem to find the smallest possible rectangular Box
> around a set of points (atoms) in space. Any suggestions or ideas
> are wellcome. I'll summerize the results.
> 
> Heiko Gerwens <gerwens@mbox.theochem.uni-hannover.de> 


Hi,

here is my answer that I believe to avoid the drawbacks mentioned by
Friedrich Ackermann <friedric@techfak.uni-bielefeld.de>

Let the original points be \vec r_i = (x_i,y_i,z_i).
For a given rotation R(\alpha,\beta,\gamma) defined via 
Euler angles \alpha, \beta, \gamma the rotated points
are \vec r_i' = R(\alpha,\beta,\gamma) \vec r_i = (x_i',y_i',z_i').

Define 

F(\alpha,\beta,\gamma) = ( max(x_i') - min(x_i') )
( max(y_i') - min(y_i') )( max(z_i') - min(z_i') )

where max and min are meant as "for all i". This is the volume of the
smallest box parallel to the coordinate axes for the rotated points.

Then the volume of the smallest box should be

min_{\alpha,\beta,\gamma} F(\alpha,\beta,\gamma)

This comes close to an algorithm already... 

Or put as a sentence: Find that frame that the minimal box 
parallel to the coordinate axes in that frame is minimal!

Obviously, one could give also a "passive" interpretation (rotate
coordinate axes instead of points ...).

Maybe some mathematician will prove that the result really is THE
smallest box. But intuitively this method should be ok.

Best regards

Herbert Homeier
---------------------------------------------------------------
Dr. Herbert H. H. Homeier
Institut fuer Physikalische und Theoretische Chemie
Universitaet Regensburg
D-93040 Regensburg, Germany
Phone: +49-941-943 4720                FAX  : +49-941-943 2305
email: na.hhomeier@na-net.ornl.gov
<A HREF="http://rchs1.uni-regensburg.de/%7Ec5008/">HOMEPAGE</A>
---------------------------------------------------------------

From owner-chemistry@ccl.net  Thu Jun 13 06:13:18 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id FAA00473; Thu, 13 Jun 1996 05:38:08 -0400 (EDT)
Received: from rulgca.LEIDENUNIV.NL  for IJZERMAN@rulgca.LeidenUniv.nl
	by bedrock.ccl.net (8.7.5/950822.1) id FAA27431; Thu, 13 Jun 1996 05:38:05 -0400 (EDT)
From: <IJZERMAN@rulgca.LeidenUniv.nl>
Received: from rulgca.LeidenUniv.nl by rulgca.LeidenUniv.nl
 (PMDF V5.0-7 #15563) id <01I5UXBHPHNK000KF9@rulgca.LeidenUniv.nl> for
 chemistry@ccl.net; Thu, 13 Jun 1996 11:38:56 +0100 (MET)
Date: Thu, 13 Jun 1996 11:38:56 +0100 (MET)
Subject: school on medicinal chemistry
To: chemistry@ccl.net
Message-id: <01I5UXBHT7EQ000KF9@rulgca.LeidenUniv.nl>
X-VMS-To: IN%"chemistry@ccl.net"
MIME-version: 1.0
Content-transfer-encoding: 7BIT



Below follows information about the Vth School on Medicinal Chemistry
organized by the Leiden/Amsterdam Center for Drug Research. Please,
do not hesitate to contact me for further information
(ijzerman@rulgca.leidenuniv.nl).

===============================================================


22-25 October 1996, Noordwijkerhout, The Netherlands

Vth LACDR SCHOOL ON MEDICINAL CHEMISTRY

The development of new drugs has become largely dependent on
a deeper understanding of human (patho)physiology. 
Nowadays, pharmacology and toxicology,
both in vitro and in vivo, are essential to exploit fundamental
knowledge for the development of drug candidates. As a conse-
quence, both disciplines are more and more applied at an early
stage of drug design to guide synthetic strategies. The course
encompasses basic and advanced aspects of lead finding,
pharmacology and toxicology to provide research chemists in the
pharmaceutical industry with the appropriate background for
their daily practice. Topics include the impact of new research
fields and techniques, such as molecular biology and molecular
modelling, on drug research, as
well as a thorough introduction in pharmacodynamics,
pharmacokinetics and molecular toxicology. New in the 5th
School will be the topic of combinatorial chemistry and
molecular diversity. In addition, three case histories will give a
flavour of chance and strategy in drug development. Essential to
the course is a workshop-like case-study ('the design of a new
drug') in which all participants are actively engaged.

Acclaims from participants to the previous schools: excellent
documentation....; case-study very useful....; very clear review of
drug research today....; lectures very good....; beyond
expectation....; intellectually most stimulating.

COURSE DIRECTION

Prof.dr. H. Timmerman, Director of Research Leiden/Amsterdam
Center for Drug Research
Head Department of Pharmacochemistry, Faculty of Chemistry,
Vrije Universiteit Amsterdam, The Netherlands.
His research programme mainly concerns the different histamine
receptors and their ligands. He is a member of the editorial
boards of several international journals and editor of the series
Pharmacochemistry Library (Elsevier) and Methods and
Principles in Medicinal Chemistry (VCH). 

Dr. A.P. IJzerman
Associate Professor, Division of Medicinal Chemistry,
Leiden-/Amsterdam Center for Drug Research, Leiden
University, The Netherlands. 
Dr. IJzerman supervises the receptor research group at the
Center, and is involved in research on purinergic receptors and
drug design. He has ample experience in modern techniques
to study ligand-receptor interactions, such as molecular
modelling/computer graphics. He is a referee to major
international journals in pharmacology, medicinal chemistry and
pharmaceutics.

REGISTRATION INFORMATION
Course location
The course will be held at the Leeuwenhorst Congress Center,
Noordwijkerhout, The Netherlands. Tel. 31 252 37 88 88.

Hotel accommodation
A number of fully equipped single rooms have been reserved at
the Congress Center.

Fee
Individuals: Hfl. 2.800,-. This includes course documentation,
mid-session refreshments, lunches, dinners and hotel
accommodation with breakfast (3 nights).
Group fee: Hfl. 2.800,- for the first participant and Hfl. 2.600,-
for the second and following participants from the same
organization.
Ph.D. students: Up to five Ph.D. students may attend the course
for a reduced fee. A statement of the supervisor acknowledging
the status of Ph.D. student should accompany the registration
form. 

Registration and payment: It is advised to forward the
registration form as soon as possible in view of the limited
course capacity of 25 participants. Confirmation of registration
will be returned upon receipt, together with an invoice for the
course fee. Registration will not be final until payment is
received.
The organization reserves the right to cancel the course should
the number of registrations be lower than 12.
Notice of cancellation, with a full refund, will be given before
September 1996.
Cancellations: Cancellations with a full refund may be made
until September 1, 1996. Cancellations between September 1 and
October 1, 1996: 50% refund. No refund is possible on
cancellations received after this date.
Substitutions may be made at any time.
Arrival: October 22, 1996, 9.30-10.00 a.m.
Course language: The official language of the course is English. 

COURSE PROGRAMME
Tuesday, October 22,
 9.30 - 10.00 Registration

=====Pharmacology=====                                                     
                                                       
10.00 - 11.00 Mathematics of receptor-ligand interaction, dr.
A.P. IJzerman, Leiden, NL
11.15 - 12.15 Receptor binding in industrial perspective, dr. M.
Tulp, Weesp, NL
13.30 - 14.30 Signal transduction and second messengers,
prof.dr. A. Bast, Amsterdam NL
14.30 - 15.30 Selection of pharmacological methods, illustrated
with antidepressants, dr. C.L.E. Broekkamp, Oss, NL
16.00 - 17.00 Molecular biology in drug design, dr. R. Leurs,
Amsterdam NL
20.30 - 21.30 Case history 'Taxol', dr. G. Gallant, Brussels, B

Wednesday, October 23,
=====Pharmacology (cont)=====

 9.00 - 10.15 Clinical pharmacology: what happens to your
molecule?, prof.dr. A.F. Cohen, Leiden, NL
10.45 - 11.30 Case study in drug design: introduction, dr. L.
Cohen, Leiden, NL
11.30 - 12.30 The concept of drug selectivity, prof.dr. H.
Timmerman, Amsterdam, NL
13.30 - 17.30 Case study in drug design, participants
20.30 - 21.30 Case history 'Omeprazole', dr. P. Lindberg,
Molndal, Sweden

Thursday, October 24,
=====Fate of drugs=====

 9.00 - 10.30 Introduction to the pharmacokinetics of drugs, dr.
W. Meuldermans, Beerse, B
11.00 - 12.00 Toxicology in industry, dr. W. Coussement,
Beerse, B
12.00 - 13.00 In vitro toxicology in drug research, dr. J.F.
Nagelkerke, Leiden, NL

=====Lead finding and optimization=====

14.30 - 15.30 Combinatorial chemistry, prof.dr. H.C.J.
Ottenheijm, Oss, NL
16.00 - 17.00 Screening of biological materials, dr. S. Wrigley,
Slough, GB
20.30 - 21.30 Case history 'Knowledge protection', dr. H.W.
Raven, Amsterdam, NL

Friday, October 25,
=====Lead finding and optimization (cont.)=====

 9.00 - 10.00 3D-databases in drug design, dr. S. van Helden,
Oss, NL
10.30 - 11.30 Molecular modeling in drug design, prof.dr. J.
Tollenaere, Beerse, B, Utrecht, NL
11.30 - 12.30 Novel targets for drug design, Dr. A.J.M. van
Oosterhout, Utrecht, NL
14.00 - 16.00 Case study in drug design: presentation,
participants


REGISTRATION FORM

Vth LACDR School on Medicinal Chemistry
22-25 October 1996

Name:
Organization: 
Position:
Address:
Postal code: City:
State/country:
Telephone: Fax:

Date of arrival:       Date of departure:

Payment [ ] cheque enclosed [ ] please charge my credit card [ ]
Master/Eurocard [ ] Visa

Card number:
 ..........................................Expiry Date:....................

Signature:                              Date: 

Return by fax or regular mail to: LACDR-Secretariat, c/o Mrs.
F.J. Velthorst, P.O.Box 9502, 2300 RA Leiden,
The Netherlands. Phone: 31 71 5274341; Fax 31 71 5274277

From smillefiori@dipchi.unict.it  Thu Jun 13 07:53:52 1996
Received: from ns.dipchi.unict.it  for smillefiori@dipchi.unict.it
	by www.ccl.net (8.7.5/950822.1) id HAA01110; Thu, 13 Jun 1996 07:12:11 -0400 (EDT)
Received: from millefiori.dipchi.unict.it ([151.97.236.168])
          by ns.dipchi.unict.it (post.office MTA v1.9.3 ID# 0-11335)
          with SMTP id AAA146 for <chemistry@www.ccl.net>;
          Thu, 13 Jun 1996 13:11:57 +0100
X-Sender: smillefiori@dipchi.unict.it
X-Mailer: Windows Eudora Version 1.4.4
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
To: chemistry@www.ccl.net
From: "Millefiori Salvatore" <smillefiori@dipchi.unict.it>
Subject: software for IR
Date: Thu, 13 Jun 1996 13:11:57 +0100
Message-ID: <19960613121157448.AAA146@millefiori.dipchi.unict.it>


Hi,

I=92m looking for a PC-software to print calculated IR spectra in continuous
envelopes of lorentzians like experimental spectra, taking and converting
discrete data input (frequency and intensity of vibrational modes).

Best regards,

                                    Andrea Alparone.

=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=
=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=
=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7

Dr. Andrea Giovanni Alparone (c/o Prof. Salvatore Millefiori)
Dottorato di Ricerca Scienze Chimiche, Universit=E0 degli Studi di Catania
Address: Dip. di Scienze Chimiche Viale A. Doria,6 - 95125 Catania, ITALY.
Phone: 39-95-330319
E-mail: smillefiori@dipchi.unict.it

=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=
=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=
=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7=A7


From owner-chemistry@ccl.net  Thu Jun 13 09:53:52 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id JAA01787; Thu, 13 Jun 1996 09:21:33 -0400 (EDT)
Received: from info.cyf-kr.edu.pl  for mbskowro@cyf-kr.edu.pl
	by bedrock.ccl.net (8.7.5/950822.1) id JAA28800; Thu, 13 Jun 1996 09:21:28 -0400 (EDT)
Received: from kinga.cyf-kr.edu.pl (kinga.cyf-kr.edu.pl [149.156.4.10]) by info.cyf-kr.edu.pl (8.7.5/8.7.5) with ESMTP id PAA17088 for <chemistry@ccl.net>; Thu, 13 Jun 1996 15:20:44 +0200 (METDST)
Received: (from mbskowro@localhost) by kinga.cyf-kr.edu.pl (8.7.5/8.7.3) id PAA18639 for chemistry@ccl.net; Thu, 13 Jun 1996 15:15:21 +0200 (METDST)
From: Marek Skowronek <mbskowro@cyf-kr.edu.pl>
Message-Id: <199606131315.PAA18639@kinga.cyf-kr.edu.pl>
Subject: Amber & Gaussian
To: chemistry@ccl.net
Date: Thu, 13 Jun 1996 15:15:20 +0200 (METDST)
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text


Dear Netters,
I would like to ask you some questions
connected with my first trials of
computations with Amber & Gaussian programs.

1. I would like to visualise gaussian data
like an electrostatic potential, density and
etc. Is there any software to visualise these
data?

2. I have some problem with missing
parameters in Amber program. Is it possible
to calculate these parameters using
Gaussian94 program? Is there any literature
about it? 

I will be very grateful if you send any
information about it.


                    Thank you in advance
                    Marek Skowronek
         

From owner-chemistry@ccl.net  Thu Jun 13 10:11:52 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id JAA01764; Thu, 13 Jun 1996 09:14:38 -0400 (EDT)
Received: from medinah.atc.ucarb.com  for jas@medinah.atc.ucarb.com
	by bedrock.ccl.net (8.7.5/950822.1) id JAA28705; Thu, 13 Jun 1996 09:14:29 -0400 (EDT)
Received: from smithja.sct.ucarb.com by medinah.atc.ucarb.com (AIX 3.2/UCB 5.64/4.03)
          id AA29690; Thu, 13 Jun 1996 09:14:28 -0400
Message-Id: <v01510102ade5bb84f15c@[140.170.208.21]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Thu, 13 Jun 1996 09:16:48 -0400
To: chemistry@ccl.net
From: jas@medinah.atc.ucarb.com (Jack Smith)
Subject: Re: CCL:Smallest rectangular box


  OK, here's my intuitive answer (i.e. no proof):

  Generate a distance matrix and find the two most distal points.  Let
these two points define the Z direction (i.e. parallel to Z-axis) of the
rectangular box.  Now generate another distance matrix where the distance
is the projected distance perpendicular to the Z direction. Find the two
most distal points and define the Y direction from the projection of their
connecting vector in a plane perpendicular to Z.  The X direction is also
now fixed - perpendicular to the Y and Z directions.  The dimensions of the
rectangular box are now determined by the most distal points in each of the
axial directions, and its center (origin) is defined by the midpoints.

  There's probably a way to prove that the six most distal points defining
the faces of the enclosing rectangular box are unique and that all the
interior points are in some sense "bound" by these six points, independent
of the rotational frame or origin.

- Jack


-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
  JACK A. SMITH                 ||  Phone:  (304) 747-5797
  Union Carbide Corp.           ||  FAX:    (304) 747-5571
  Catalyst Skill Center         ||  E-mail: ajassa1@peabody.sct.ucarb.com
  P.O. Box 8361                 ||          jas@medinah.atc.ucarb.com **
  S. Charleston, WV  25303      ||          ** Expiring very soon
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=



From Darren.Andrews@man.ac.uk  Thu Jun 13 10:53:51 1996
Received: from nessie.mcc.ac.uk  for Darren.Andrews@man.ac.uk
	by www.ccl.net (8.7.5/950822.1) id KAA02255; Thu, 13 Jun 1996 10:50:46 -0400 (EDT)
Received: from [130.88.12.235] (actually femt03.ch.man.ac.uk) 
          by nessie.mcc.ac.uk with SMTP (PP); Thu, 13 Jun 1996 15:50:13 +0100
X-Sender: mbcdpa@130.88.200.20
Message-Id: <v02140b02ade5da789993@[130.88.12.235]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Thu, 13 Jun 1996 15:51:02 +0100
To: chemistry@www.ccl.net
From: Darren.Andrews@man.ac.uk (Darren Andrews)
Subject: MD simulations of liquids and electron transfer


I am looking for a program that will calculate molecular dynamics for
solvent/solute systems where a reaction (electron transfer) occurs.  Is
there a general code for the liquid dynamics bit and is there a standard
way of modelling the reaction?

Cheers.


Darren Andrews.

University of Manchester,
Department of Chemistry,
Oxford road,
Manchester.
M13 9PL.
England.

Darren.Andrews@man.ac.uk
PGP: 512/62A971A9

Tel: 0161 275  2000 ext.4699.
Fax: 0161 275 4598.



From owner-chemistry@ccl.net  Thu Jun 13 10:59:20 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id KAA02166; Thu, 13 Jun 1996 10:41:41 -0400 (EDT)
Received: from punch.ic.ac.uk  for p.harvey01@ic.ac.uk
	by bedrock.ccl.net (8.7.5/950822.1) id KAA00068; Thu, 13 Jun 1996 10:41:40 -0400 (EDT)
Received: from judy.ic.ac.uk by punch.ic.ac.uk with SMTP (PP);
          Thu, 13 Jun 1996 15:41:29 +0100
Received: from chcmga.ch.ic.ac.uk (chcmga-fddi.ch.ic.ac.uk [155.198.2.33]) 
          by judy.ic.ac.uk (8.7.5/8.7.5) with SMTP id PAA20428 
          for <chemistry@ccl.net>; Thu, 13 Jun 1996 15:40:50 +0100 (BST)
Received: from chcgc.ch.ic.ac.uk 
          by chcmga.ch.ic.ac.uk (940816.SGI.8.6.9/4.0)          id PAA17286;
          Thu, 13 Jun 1996 15:40:50 +0100
X-Received: from localhost by chcgc.ch.ic.ac.uk; Thu, 13 Jun 1996 15:40:49 
            +0100
Date: Thu, 13 Jun 1996 15:40:48 +0100 (bst)
From: Pat Harvey <p.harvey01@ic.ac.uk>
X-Sender: pharve@chcgc
To: chemistry@ccl.net
Subject: Re: CCL:Smallest rectangular box
In-Reply-To: <v01510102ade5bb84f15c@[140.170.208.21]>
Message-ID: <Pine.SGI.3.93.960613153346.25021D-100000@chcgc>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


On Thu, 13 Jun 1996, Jack Smith wrote:

>   OK, here's my intuitive answer (i.e. no proof):
> 
>   Generate a distance matrix and find the two most distal points.  Let
> these two points define the Z direction (i.e. parallel to Z-axis) of the
> rectangular box.  Now generate another distance matrix where the distance
> is the projected distance perpendicular to the Z direction. Find the two
> most distal points and define the Y direction from the projection of their
> connecting vector in a plane perpendicular to Z.  The X direction is also
> now fixed - perpendicular to the Y and Z directions.  The dimensions of the
> rectangular box are now determined by the most distal points in each of the
> axial directions, and its center (origin) is defined by the midpoints.
> 
>   There's probably a way to prove that the six most distal points defining
> the faces of the enclosing rectangular box are unique and that all the
> interior points are in some sense "bound" by these six points, independent
> of the rotational frame or origin.
> 
> - Jack
> 
> 
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>   JACK A. SMITH                 ||  Phone:  (304) 747-5797
>   Union Carbide Corp.           ||  FAX:    (304) 747-5571
>   Catalyst Skill Center         ||  E-mail: ajassa1@peabody.sct.ucarb.com
>   P.O. Box 8361                 ||          jas@medinah.atc.ucarb.com **
>   S. Charleston, WV  25303      ||          ** Expiring very soon
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

Surely this is not correct. Consider the trivial case of two points. The
smallest bounding rectangle will have the two points at opposite corners.

Pat Harvey                  Tel:   +44-71-594-5847                           
New building Room 164 
Department of Chemistry     email: pharve@ic.ac.uk                         
Imperial College                                                           
London. SW7 2AY             Grad Student           




From owner-chemistry@ccl.net  Thu Jun 13 11:53:54 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id LAA02389; Thu, 13 Jun 1996 11:09:57 -0400 (EDT)
Received: from punch.ic.ac.uk  for p.harvey01@ic.ac.uk
	by bedrock.ccl.net (8.7.5/950822.1) id LAA00595; Thu, 13 Jun 1996 11:09:56 -0400 (EDT)
Received: from judy.ic.ac.uk by punch.ic.ac.uk with SMTP (PP);
          Thu, 13 Jun 1996 16:09:41 +0100
Received: from chcmga.ch.ic.ac.uk (chcmga-fddi.ch.ic.ac.uk [155.198.2.33]) 
          by judy.ic.ac.uk (8.7.5/8.7.5) with SMTP id QAA21652 
          for <chemistry@ccl.net>; Thu, 13 Jun 1996 16:08:52 +0100 (BST)
Received: from chcgc.ch.ic.ac.uk 
          by chcmga.ch.ic.ac.uk (940816.SGI.8.6.9/4.0)          id QAA18225;
          Thu, 13 Jun 1996 16:08:51 +0100
X-Received: from localhost by chcgc.ch.ic.ac.uk; Thu, 13 Jun 1996 16:08:50 
            +0100
Date: Thu, 13 Jun 1996 16:08:50 +0100 (bst)
From: Pat Harvey <p.harvey01@ic.ac.uk>
X-Sender: pharve@chcgc
To: chemistry@ccl.net
Subject: Re: Smallest rectangular box - correction
In-Reply-To: <v01510102ade5bb84f15c@[140.170.208.21]>
Message-ID: <Pine.SGI.3.93.960613160521.25915B-100000@chcgc>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


On Thu, 13 Jun 1996, Jack Smith wrote:

>   OK, here's my intuitive answer (i.e. no proof):
> 
>   Generate a distance matrix and find the two most distal points.  Let
> these two points define the Z direction (i.e. parallel to Z-axis) of the
> rectangular box.  Now generate another distance matrix where the distance
> is the projected distance perpendicular to the Z direction. Find the two
> most distal points and define the Y direction from the projection of their
> connecting vector in a plane perpendicular to Z.  The X direction is also
> now fixed - perpendicular to the Y and Z directions.  The dimensions of the
> rectangular box are now determined by the most distal points in each of the
> axial directions, and its center (origin) is defined by the midpoints.
> 
>   There's probably a way to prove that the six most distal points defining
> the faces of the enclosing rectangular box are unique and that all the
> interior points are in some sense "bound" by these six points, independent
> of the rotational frame or origin.
> 
> - Jack
> 
> 
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>   JACK A. SMITH                 ||  Phone:  (304) 747-5797
>   Union Carbide Corp.           ||  FAX:    (304) 747-5571
>   Catalyst Skill Center         ||  E-mail: ajassa1@peabody.sct.ucarb.com
>   P.O. Box 8361                 ||          jas@medinah.atc.ucarb.com **
>   S. Charleston, WV  25303      ||          ** Expiring very soon
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> 
Consider 4 point defining a rectangle. The smallest bounding rectangle is
the rectangle defined by the points. The most distal points are the
diagonals.

Pat Harvey                  Tel:   +44-71-594-5847                           
New building Room 164 
Department of Chemistry     email: pharve@ic.ac.uk                         
Imperial College                                                           
London. SW7 2AY             Grad Student          




From owner-chemistry@ccl.net  Thu Jun 13 12:53:54 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id MAA03263; Thu, 13 Jun 1996 12:33:53 -0400 (EDT)
Received: from medinah.atc.ucarb.com  for jas@medinah.atc.ucarb.com
	by bedrock.ccl.net (8.7.5/950822.1) id MAA02191; Thu, 13 Jun 1996 12:33:36 -0400 (EDT)
Received: from smithja.sct.ucarb.com by medinah.atc.ucarb.com (AIX 3.2/UCB 5.64/4.03)
          id AA28181; Thu, 13 Jun 1996 12:33:32 -0400
Message-Id: <v01510105ade5f209b571@[140.170.208.21]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Thu, 13 Jun 1996 12:35:49 -0400
To: chemistry@ccl.net
From: jas@medinah.atc.ucarb.com (Jack Smith)
Subject: Re: CCL:Smallest rectangular box


>
>Surely this is not correct. Consider the trivial case of two points. The
>smallest bounding rectangle will have the two points at opposite corners.
>

  Actually, the bounding rectangle for two points would be an infinitely
thin rectangle with the two points at either end. The goal is to minimize
the volume (area), not its largest dimension.  The smallest bounding square
would indeed have the points in opposite corners.

- Jack



From heather@xhost1.tripos.com  Thu Jun 13 13:02:53 1996
Received: from gatekeeper.tripos.com  for heather@xhost1.tripos.com
	by www.ccl.net (8.7.5/950822.1) id MAA03154; Thu, 13 Jun 1996 12:12:03 -0400 (EDT)
Received: (from news@localhost) by gatekeeper.tripos.com (8.6.12/8.6.12) id LAA19983 for <CHEMISTRY@www.ccl.net>; Thu, 13 Jun 1996 11:12:17 -0500
Received: from tripos.tripos.com(192.160.145.1) by gatekeeper.tripos.com via smap (V1.3)
	id sma019981; Thu Jun 13 11:11:57 1996
Received: from xhost1.tripos (xhost1.tripos.com [192.160.145.212]) by tripos.tripos.com (8.6.12/8.6.12) with ESMTP id LAA01721 for <CHEMISTRY@www.ccl.net>; Thu, 13 Jun 1996 11:09:15 -0500
Received: by xhost1.tripos (940816.SGI.8.6.9/5.6)
	id LAA03354; Thu, 13 Jun 1996 11:09:17 -0500
Date: Thu, 13 Jun 1996 11:09:17 -0500
From: heather@xhost1.tripos.com (Heather Hunter)
Message-Id: <199606131609.LAA03354@xhost1.tripos>
To: CHEMISTRY@www.ccl.net
Subject: Seminars



A final notice of the free Alchemy 2000 seminar tour, after much sucess
in Chicago, Boston and New Jersey.  The Alchemy 2000 tour continues
next week in California.  Hope to see you there

Chemical Discovery On Your PC.
See the power of the new scientific technology for the desktop computer!
Locations

LaJolla, CA			June 18
Redwood City, CA		June 19

You are invited to a FREE half-day workshop on computational applications for chemical discovery with the new integrated Alchemy 2000 chemical discovery system.  Learn how you can use your current chemistry software with this new, advanced system for chem

Discovery new personal computing technologies for:
	¥	chemical property prediction
	¥	molecular structure analysis
	¥	accurate energy calculations
	¥	advanced graphics

Alchemy 2000Õs optional add-on modules for special applications such as QSAR, polymer prediction, protein analysis and LogP will be discussed with examples and demonstrations.

Register now for the Alchemy 2000 Preview!

Name
Affiliation
Address
City
State
Zip
Tel
Fax
Email
Indicate which preview session you will attend:
Location			Date

DEADLINE IS ONE WEEK PRIOR TO YOUR CHOICE OF PREVIEW DATE

Call, fax, or email us to register or find out more.  Refreshments and additional application materials will be provided.
Tel	1-800-323-2960, Ext. 3242	email mcain@tripos.com
Web	Tripos Home Page		Fax  Tripos, Inc. 314-647-9241
          URL: http://www.tripos.com
          [Support, Events, Alchemy 2000 Tour)

Conformation and location details will be sent upon receipt of your registration.

Thank you
Heather

From owner-chemistry@ccl.net  Thu Jun 13 13:07:17 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id MAA03143; Thu, 13 Jun 1996 12:10:38 -0400 (EDT)
Received: from msi.com  for ryszard@msi.com
	by bedrock.ccl.net (8.7.5/950822.1) id MAA01840; Thu, 13 Jun 1996 12:10:37 -0400 (EDT)
Received: from aga.msi.com by msi.com (4.1/SMI-4.1)
	id AA00641; Thu, 13 Jun 96 11:59:35 EDT
Received: by aga.msi.com (5.65/DEC-OSF/1-29Feb92);
	id AA04870; Thu, 13 Jun 1996 12:10:22 -0400
Date: Thu, 13 Jun 1996 12:10:22 -0400 (EDT)
From: Ryszard Czerminski <ryszard@msi.com>
To: Pat Harvey <p.harvey01@ic.ac.uk>
Cc: chemistry@ccl.net
Subject: Re: CCL:Smallest rectangular box
In-Reply-To: <Pine.SGI.3.93.960613153346.25021D-100000@chcgc>
Message-Id: <Pine.OSF.3.91.960613120705.4626B-100000@aga>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


I do not quite understand this:

> Surely this is not correct. Consider the trivial case of two points. The
> smallest bounding rectangle will have the two points at opposite corners.

but if you consider an example of square with
side lenght a, then procedure below will give you as a bounding box
square with side lenght = a*sqrt(2)

Best regards,

Ryszard

 ..................................
Ryszard Czerminski
Molecular Simulations Incorporated
16 New England Executive Park
Burlington, MA 01803-5297
tel: (617)229-8875 x 217
fax: (617)229-9899
e-mail: ryszard@msi.com

On Thu, 13 Jun 1996, Pat Harvey wrote:

> On Thu, 13 Jun 1996, Jack Smith wrote:
> 
> >   OK, here's my intuitive answer (i.e. no proof):
> > 
> >   Generate a distance matrix and find the two most distal points.  Let
> > these two points define the Z direction (i.e. parallel to Z-axis) of the
> > rectangular box.  Now generate another distance matrix where the distance
> > is the projected distance perpendicular to the Z direction. Find the two
> > most distal points and define the Y direction from the projection of their
> > connecting vector in a plane perpendicular to Z.  The X direction is also
> > now fixed - perpendicular to the Y and Z directions.  The dimensions of the
> > rectangular box are now determined by the most distal points in each of the
> > axial directions, and its center (origin) is defined by the midpoints.
> > 
> >   There's probably a way to prove that the six most distal points defining
> > the faces of the enclosing rectangular box are unique and that all the
> > interior points are in some sense "bound" by these six points, independent
> > of the rotational frame or origin.
> > 
> > - Jack
> > 
> > 
> > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> >   JACK A. SMITH                 ||  Phone:  (304) 747-5797
> >   Union Carbide Corp.           ||  FAX:    (304) 747-5571
> >   Catalyst Skill Center         ||  E-mail: ajassa1@peabody.sct.ucarb.com
> >   P.O. Box 8361                 ||          jas@medinah.atc.ucarb.com **
> >   S. Charleston, WV  25303      ||          ** Expiring very soon
> > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> 
> Surely this is not correct. Consider the trivial case of two points. The
> smallest bounding rectangle will have the two points at opposite corners.
> 
> Pat Harvey                  Tel:   +44-71-594-5847                           
> New building Room 164 
> Department of Chemistry     email: pharve@ic.ac.uk                         
> Imperial College                                                           
> London. SW7 2AY             Grad Student           
> 
> 
> 
> 
> -------This is added Automatically by the Software--------
> -- Original Sender Envelope Address: owner-chemistry@ccl.net
> -- Original Sender From: Address: p.harvey01@ic.ac.uk
> CHEMISTRY@www.ccl.net: Everybody | CHEMISTRY-REQUEST@www.ccl.net: Coordinator
> MAILSERV@www.ccl.net: HELP CHEMISTRY or HELP SEARCH | Gopher: www.ccl.net 73
> Anon. ftp: www.ccl.net   | CHEMISTRY-SEARCH@www.ccl.net -- archive search
>              Web: http://www.ccl.net/chemistry.html 
> 
> 

From owner-chemistry@ccl.net  Thu Jun 13 13:53:55 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id NAA03447; Thu, 13 Jun 1996 13:06:03 -0400 (EDT)
Received: from bioc1.msi.com  for marvin@bioc1.msi.com
	by bedrock.ccl.net (8.7.5/950822.1) id NAA02658; Thu, 13 Jun 1996 13:05:59 -0400 (EDT)
Received: from iris147.biosym.com by bioc1.msi.com (5.64/0.0)
	id AA07934; Thu, 13 Jun 96 10:04:48 -0700
Received: by iris147.msi.com (940816.SGI.8.6.9/930416.SGI)
	for chemistry@ccl.net id KAA06565; Thu, 13 Jun 1996 10:05:24 -0700
Date: Thu, 13 Jun 1996 10:05:24 -0700
From: marvin@msi.com (Marvin Waldman)
Message-Id: <199606131705.KAA06565@iris147.msi.com>
To: chemistry@ccl.net
Subject: RE:SMALLEST BOUNDING BOX


Since this thread will not seem to die, I thought I would throw in my own 
observations about the subject.  Using principal axes is probably a 
reasonable approximation for most molecules, but is certainly not the
full analytical solution.  Implementing some kind of optimization procedure
(genetic algorithm, simplex, etc.) to calculate the bounding box for
different orientations of the molecule is also quite reasonable, but
is not guaranteed to find a global minimum solution (local minima are possible).

A correct analytical procedure can be written fairly straightforwardly and
involves determining something called the convex hull for the molecule
by generating planes corresponding to all triplets of points and then
finding the distance from each of the remaining points to each of the planes.
Any planes for which there are points on both sides of the plane can
be immediately rejected.  The remaining planes form the convex hull of
the molecule.  For each of these planes, determine the maximum distance
to all remaining points.  Take the point maximally distant to the plane,
and generate a parallel plane through it.  Project all the remaining
points onto one of the parallel planes.  Now the problem is reduced
to finding the bounding 2 dimensional rectangle for these projected
points.  This can be done by finding all pairs of lines through the
points.  Reject any lines for which there are points on both sides
of the line.  The remaining lines form the 2 dimensional hull for
these points.  For each such line, find the maximally distance point.
Draw a line through this point parallel to its corresponding line.
Now project all the remaining points onto one of these lines.
Find the two points with largest extent along this line.  Draw a
pair of lines through them perpendicular to the original line.  Now take
the 4 lines (which generate the bounding rectangle in the plane) and
generate planes through them perpendicular to the original plane.  The
6 planes generate a bounding box for the original points.  After generating
all such possible bounding boxes (considering all planes of the 3D
convex hull for the molecule and all lines for the 2D convex hull for the 
projected points in each of these planes), choose the bounding box with the
smallest volume (or surface area or whatever other criterion you may
choose to apply).  

The above algorithm while correct and fully analytical is not 
optimal.  There are many ways to improve the efficiency of this procedure.
The problem has been investigated at length in the computer math literature.
A good discussion may be found in:

"Putting objects into boxes", by R.R. Martin and P.C. Stephenson in
Computer Aided Design, Volume 20, pp. 506-514, (1988).

By the way, the above only discusses a box enclosing the points which
I assume correspond to the atoms of the molecule.  Possibly a more
relevant application to molecular modeling is the minimum bounding
box enclosing the VDW ( dare I say v(V?)an der Waals('?) ) surface
of the molecule.  This has a different (although related) solution
to the above problem.

Regards,

Marvin Waldman, Ph.D.
Director, Rational Drug Design
Molecular Simulations Inc.
e-mail: marvin@msi.com
Web:    http://www.msi.com

From owner-chemistry@ccl.net  Thu Jun 13 14:53:54 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id OAA04079; Thu, 13 Jun 1996 14:35:11 -0400 (EDT)
Received: from medinah.atc.ucarb.com  for jas@medinah.atc.ucarb.com
	by bedrock.ccl.net (8.7.5/950822.1) id OAA03946; Thu, 13 Jun 1996 14:35:01 -0400 (EDT)
Received: from smithja.sct.ucarb.com by medinah.atc.ucarb.com (AIX 3.2/UCB 5.64/4.03)
          id AA30433; Thu, 13 Jun 1996 14:34:54 -0400
Message-Id: <v01510106ade609923b41@[140.170.208.21]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Thu, 13 Jun 1996 14:37:14 -0400
To: chemistry@ccl.net
From: jas@medinah.atc.ucarb.com (Jack Smith)
Subject: Re: CCL:Smallest rectangular box - correction


>Consider 4 point defining a rectangle. The smallest bounding rectangle is
>the rectangle defined by the points. The most distal points are the
>diagonals.

   Hmmm. I believe you're right.  And it seems to translates to the 3-D
case as well.  So much for intuition.  I think you and Dominic are both on
the right track regarding the use of the "diagonals" instead of the axial
dimensions.

   Well, I guess it's time for someone who actually knows the answer to
chime in and finally eliminate this "noise" from the CCL, or at least take
the discussion offline for a while and post a summary.

- Jack



From Y0H8797@ACS.TAMU.EDU  Thu Jun 13 15:53:55 1996
Received: from VMS1.TAMU.EDU  for Y0H8797@ACS.TAMU.EDU
	by www.ccl.net (8.7.5/950822.1) id OAA04212; Thu, 13 Jun 1996 14:55:41 -0400 (EDT)
Date: Thu, 13 Jun 1996 13:55:12 -0500 (CDT)
From: YONG HUANG <Y0H8797@ACS.TAMU.EDU>
To: CHEMISTRY@www.ccl.net
Message-Id: <960613135512.20a469c7@ACS.TAMU.EDU>
Subject: A MOPAC Error Message; Geometry Optimize at Excited Singlet State


  I'm using CAChe MOPAC 94 on a Mac. From time to time I get an error message 
  in the output file upon finishing a geometry optimization
  
  "Trust radius now less than 0.00100 Optimization terminating
    Geometry may not be completely optimized"
  
  What does it mean? Other than this message, everything is fine except 
  sometimes the gradient accompanying the final heat of formation is too big 
  (>5 e.g.). (Only simetimes though). I use EF, ANALYT, PRECISE, PM3, etc. 
  keywords. Thanks for help.
  
  Another thing I'd like to say. I used to have problems geometry optimizing a
  molecule (a typical size like C10H10O10) in its excited singlet state. Got 
  help from CCL and used keyword NOANCI. Worked for only some of my molecules.
  Besides, NOANCI tended to give me large final gradient (>1 if I remember
  right). Now I find that if I run a CI at excited singlet state and then
  geometry optimize it, it works much better. Run CI at as high levels as you
  and/or computer can tolerate. [So far I have tested C10H10O10 only since each
  test takes many hours. But I feel this "CI->GeomOpt" method is generally
  applicable.] Any comments are welcome.
  
  Yong
  y0h8797@acs.tamu.edu

From cory@bohr.chem.mun.ca  Thu Jun 13 16:54:02 1996
Received: from piva.ucs.mun.ca  for cory@bohr.chem.mun.ca
	by www.ccl.net (8.7.5/950822.1) id QAA04506; Thu, 13 Jun 1996 16:13:34 -0400 (EDT)
Received: from bohr.chem.mun.ca (bohr.chem.mun.ca [134.153.14.2]) by piva.ucs.mun.ca (8.7.5/8.7.3) with SMTP id RAA03757 for <@morgan.ucs.mun.ca:chemistry@www.ccl.net>; Thu, 13 Jun 1996 17:43:34 -0230 (NDT)
Received: by bohr.chem.mun.ca (931110.SGI/921111.SGI.AUTO)
	for @morgan.ucs.mun.ca:chemistry@www.ccl.net id AA27387; Thu, 13 Jun 96 17:46:54 -0230
Date: Thu, 13 Jun 96 17:46:54 -0230
From: cory@bohr.chem.mun.ca (Cory C. Pye)
Message-Id: <9606132016.AA27387@bohr.chem.mun.ca>
To: chemistry@www.ccl.net
Subject: Smallest rectangular box


To solve the bounding box problem:
Step 1: Metric:
Decide on a metric to define smallest. One that works well is the volume of  
the box. V(d1,d2,d3)=d1*d2*d3. Another would be that the largest dimension of 
the box be as small as possible which would imply that (without loss of 
generality) V(d1,d2,d3)=d1 subject to d1>d2>d3 . We use the former.
(It is implicit that d1,d2,d3 >= 0)

Step 2: Define your constraints: 
Linear Algebra: Let n_i be the unit normal vectors of the 3 families 
of planes defining the possible boxes. A plane containing a boundary would 
be generally expressed as <n1 , (P-P0)> = 0, (where <,>=dot product,
P0 is any chosen point in the plane and P is any other point in the 
plane). The line given by P(k1)=k1*n1 (passing through the origin) must intersect
the plane, and we choose this point to be P0, since 
<n1, k1*n1>=k1*<n1,n1>=k1*1=k1 (i.e. the plane is k1 units away from the origin).
This choice of P0 allows us to define the two parallel planes as 
<n1,P>=k1
<n1,P>=k1+d1 
where d1 is one of the dimension of the box. 
Similarly,
<n2,P>=k2
<n2,P>=k2+d2
and 
<n3,P>=k3
<n3,P>=k3+d3
With the proviso that <n1,n2>=<n1,n3>=<n2,n3>=0, 
which ensures that the planes are mutually perpendicular, the constraints are
If we have N points x1,...,xn, for each xi,
k1 < <n1,xi>  < k1+d1
k2 < <n2,xi>  < k2+d2
k3 < <n3,xi>  < k3+d3
giving 6*N inequality constraints and 3 equality constraints.
We note that the matrix formed by N=[n1,n2,n3] is orthogonal, i.e. N^{-1}=N^{t}.

Thus the problem can be restated as,
for a collection of N points x_i, find an orthogonal matrix N and 2 vectors
k = [k1:k2:k3]^{T} and d = [d1:d2:d3]^{T} such that d1*d2*d3 is minimized subject
to the conditions:

 0 < N*x_i - k < d (componentwise)

N corresponds geometrically to a rotation and k to a translation matrix of the 
points.

The problem is solved for less than 4 points, since for three points, 
the dimension normal to the plane defined by these points is zero, for two 
points, any direction perpendicular to the line, and for one point, any 
dimension.

Discussion of other postings:
This makes the point by p.harvey01@ic.ac.uk irrelevant since the `opposite 
corners' are equivalent to the opposite sides (case N=2).

>Surely this is not correct. Consider the trivial case of two points. The
>smallest bounding rectangle will have the two points at opposite corners.

The point by friedric@techfak.uni-bielefeld.de is valid. The convex hull of the
set would have to be enclosed by this box, and could reduce the dimensionality
of the problem considerably, since only the points on the hull would need to be
considered in the constraints. The faces are probably not related however, since 
the likelihood that any two faces opposite to each other would be parallel is 
small, and given any face A, B, C, and the point P farthest from the plane 
containing it, a slight rotation of the normal of the plane toward avg(A,B,C)-P
would shorten the distance from the plane to P, and thus the most likely 
scenario would entail a vertex pair as being on the two planes.

>to get an idea of  a correct answer.   I  think  the  faces of the
>rectangular  box  are  related to the faces of the boundary of the
>convex hull of the point set.

Many of the points relating to principal axes would not apply because V is not
quadratic.

A nonlinear optimization method with constraints may be needed to solve this 
problem, which might not have an analytic solution determinable a priori.
Using the second metric would probably have an analytic solution relating to
the principal axes, since the problem is then quadratic.
   *************
 *****************  !  Cory C. Pye
***   **    **  **  !  Graduate Student and Unpaid Sys Admin
**   *  ****        !  Theoretical and Computational Chemistry
**      *  *        !  cory@bohr.chem.mun.ca
**      *  *        !  http://www.ucs.mun.ca/~cory/index.html
***     *  *    **  !
 *****************  !  Les Hartree-Focks
   *************    !  (Apologies to Montreal Canadien Fans)



From owner-chemistry@ccl.net  Thu Jun 13 17:53:56 1996
Received: from bedrock.ccl.net  for owner-chemistry@ccl.net
	by www.ccl.net (8.7.5/950822.1) id QAA04748; Thu, 13 Jun 1996 16:58:25 -0400 (EDT)
Received: from SIUCVMB.SIU.EDU  for TAPASKAR@SIUCVMB.SIU.EDU
	by bedrock.ccl.net (8.7.5/950822.1) id QAA05675; Thu, 13 Jun 1996 16:58:25 -0400 (EDT)
Message-Id: <199606132058.QAA05675@bedrock.ccl.net>
Received: from SIUCVMB.SIU.EDU by SIUCVMB.SIU.EDU (IBM VM SMTP V2R2)
   with BSMTP id 3018; Thu, 13 Jun 96 15:58:04 CST
Date: Thu, 13 Jun 96 15:57:11 CST
From: "Tapas Kar" <TAPASKAR@SIUCVMB.SIU.EDU>
To: chemistry@ccl.net
Subject: Borazine


We are presently working on an ab initio project examining borazine with
-OH, -NO2, -NH2, and -CH3 substituents.  In the literature, we have found a
few articles (theoretical studies) with such substitutions on the borazine
ring, but all of these substitutions have been on the B atom of the ring.
Why haven't there been any studies in which the substituents are on the N
atoms?  Would a substitution or addition reaction at the N atom be possible
for a borazine ring?  And, has anyone performed such a reaction?

Thanks!

Tapas Kar and Don Elmore