Reasoning," in "Computers in Chemical Education and Research," E.V.
Ludena, N.H. Sabelli, amd A.C. Wahl, Eds., Plenum Press, New York,
N.Y., 1977, p. 461.

(6) D.H. Smith and R.E. Carhart, "Structure Elucidation Based on
Computer Analysis of High and Low Resolution Mass Spectral Data," in
"High Performance Mass Spectrometry: Chemical Applications,” M.L.
Gross, &d., American Chemical Society, 1978, p. 325.

(7) T.H. Varkony, D.H. Smith, and C. Djerassi, "Computer-Assisted
Structure Manipulation: Studies in the Biosynthesis of Natural
Products," Tetrahedron, 34, 841 (1978).

(8) D.H. Smith and P.c. Jurs, "Prediction of 13C NMR Chemical
Shifts," J. Am. Chem. Soc., 100, 3316 (1978).

(9) T.H. Varkony, R.E. Carhart, D.H. Gmith, and C. Djerassi,
"Computer-Assisted Simulation of Chemical Reaction Sequences.
Applications to Problems of Structure Elucidation," J. Chem. Inf. Comp.
Sci., 18, 168 (1978). —

(18) D.H. Gnith, T.C. Rindfleisch, and W.J. Yeager, "Exchange of
Comments: Analysis of Complex Volatile Mixtures by a Combined Gas
Chromatography-Mass Spectrometry System," Anal. Chem., 58, 1585 (1978).

(11) W.L. Fitch, P.J. Anderson, and” D.H. Smith, “Isolation,
Identification and Quantitation of Urinary Organic Acids," J. Chrom.,
in press.

(12) W.L. Pitch, E.T. Everhart, and D.H. Smith, "Characterization
of Carbon Black Adsorbates and Artifacts Formed During Extraction,"
Anal. Chem., in press,

 

(13) W.L. Fitch and D.H. Smith, "Analysis of Adsorption Properties
and Adsorbed Species on Cammercial Polymeric Carbons," Environ. Sci.
Tech., in press.

(14) J.G. Nourse, R.E. Carhart, D.H. Smith, and C. Djerassi,
"Exhaustive Generation of Stereoisomers for Structure Elucidation," J.
Am. Chem. Soc., in press. .

(15) C. Djerassi, D.H. Smith, and T.H. Varkony, "A Novel Role of
Computers in the Natural Products Field,” Naturwiss., in press.

(16) N.A.B. Gray, D.H. Smith, T.8. Varkony, R.E. Carhart, and B.G.
Buchanan, "Use of a Computer to Identify Unknown Compounds. The
Automation of Scientific Inference," Chapter 7 in "Biomedical
Applications of Mass Spectrometry," G.R. Waller, Ed., in press.

126
(17) T.C. Rindfleisch and D.H. Smith, in Chapter 3 of "Biomedical
Applications of Mass Spectrometry," G.R. Waller, Ed., in press.

(18) T.H. Varkony, Y. Shiloach, and D.H. Gnith, "Computer-Assisted
Examination of Chemical Compounds for Structural Similarities," J.
Chem. Inf. Comp. Sci., in press.

(19) R. Carlson, et al., Bioorg. Chem., 7, in press.

(20) J.G. Nourse, “The Configuration Symmetry Group and its
Application to Stereoisomer Generation, Specification and Enumeration,"
J. Am. Chem. Soc., in press.

3 SIGNIFICANCE

There are several results obtained during the past year which are
especially significant for both the advancement of our techniques for
biomolecular structure elucidation and the dissemination of results of
our research. Perhaps the most significant result is in the category
of dissemination because for the first time we can offer directly to
the biomedical community the results of our research in the form of
exportable computer software for computer-assisted structure
elucidation. We have discussed in the past the differences between our
work and many other forms of health-related research. The latter can
generally be described in the literature in sufficient detail for
others to duplicate the procedures and results and, more importantly,
build on those results in extensions of the work. When one of the
primary “results” of research is a set of complex computer programs,
transfer of these results becomes a formidable task. Certainly the
programs and even some of the algorithms can be described in the
literature. We have done so, and the recent references at the end of
the previous section represent such publications. However, unless an
interested person can actually gain access to a program in an
executable, or runnable, form, such descriptions are inadequate. We
now have the capability of offering the new, smaller, faster version of
CONGEN to the community at SUMEX on a trial basis and exported to their
own computers for those who wish to make more extensive use of the
program. Therefore, we can share our results with our collaborators
much more easily now than in the past and persons receiving the program
nave a foundation on which to build.

The workshops were also significant in that we received
constructive suggestions and criticisms from a _ cross-section of
potential users of our computational techniques, while at the same time
exposing a variety of persons active in structure elucidation to CONGEN
working on real problems. The success of this effort has encouraged us
to make available in the exportable CONGEN a variety of other structure
Manipulation tools which we and the persons at the workshop perceive as
useful as adjuncts to CONGEN. ‘Thus, as we have developed capabilities
for exploration of large numbers of structures in the STRUCC program,
and new ways to use and display the results of the STEREO program, we
have begun incorporating these capabilities into CONGEN. ‘This effort
will continue, using CONGEN as the focal point for further developments
in the area (see next section).

Other significant advances have been summarized in the sections
outlining various aspects of our research, above. These include new
mathematical developments which made possible the reduced size and
increased efficiency of CONGEN, completion of the STEREO program and
integration of it into CONGEN, new approaches to rule formation in
Meta-DENDRAL and better methods for predicting mass spectra and ranking
Structures based on those predictions. The last item deserves further
comment, because it represents a general approach to structural
analysis of which we can now take advantage because of the efficiency
o£ the new CONGEN. Spectrum prediction for a variety of spectroscopic
techniques, including mass, NMR, IR, etc., including now even
chiroptical methods for future work, represents a valuable method for
structure determination. Now that we can deal with hundreds or even
thousands of structures in intermediate stages of a problem, such
Spectrum prediction techniques provide powerful filters to separate
plausible from implausible structures.

4 RESEARCH GOALS 1979-1986

We have several goals for CONGEN development and export during the
next grant year. Because several were summarized previously, we give
only a brief list here. The reprogramming effort will continue, first
to incorporate aromaticity into the program and second to include
capabilities for constraints interpretation and translation. These
capabilities have been experimental efforts in the old, LISP version of
CONGEN. Some additional development must be done and the BCPL version
brought up to date with these features. Such developments will allow
much more intuitive use of the program (see section on Conclusions from
the Workshops) and greatly improve the ease with which structural
problems can be specified to CONGEN.

We will be pushing very hard for further export of CONGEN either
by developments here at Stanford or by assisting others at remote
sites. We can now supply version for DEC-19 and DEC-29 under TOPS-19
or 28 or TENEX operating systems. The contract for an NIH/EPA CIS
version will make CONGEN available on that system soon. Our Joint
Study project will hopefully allow us to develop an IBM version at

128
minimal cost. A version for the CDC systems at the National Resource
for Computation in Chemistry is umder study. The whole area of mini-
computer versions is under investigation. We expect that these efforts
will make the current version widely available. However, it does leave
open the question of long-term maintenance and, particularly, upgrading
older versions as new developments ensue. These issues will be dealt
with at length in our renewal proposal for funding subsequent to the
coming year,

In further response to the workshop requests, we will be
developing an extended "help" facility for the program which, together
with improved documentation, will improve the ease with which new
persons can become familiar with CONGEN. We have already begun
investigations of how to improve the current structure drawing
facilities in CONGEN, with our first goal to improve the teletype
drawings so that all users can benefit. We will also be considering
methods for aiding those with graphics terminals to exploit the
improved structure input and output possible with such devices.

Further work on the STEREO program will be to develop further the
constrained stereoisomer generator and to improve interaction with the
main CONGEN program and the user. Specifically, constraints involving
the chirality of structures, patterns of equivalent atoms based on
either stereoisomer or topological symmetry, and undesirable structural
features such as trans double bonds in small rings will be implemented.
The flow of information within the program will became two-way between
CONGEN and the stereoisomer generator. At present, this information
only flows to the stereoisomer generator. The user interaction will be
improved so the user can more easily visualize the stereoisomers and
the stereochemistry of substructures and can input stereochemical
information more intuitively.

We will emphasize several lines of investigation with regards to
examination and evaluation of structural candidates from CONGEN,
through the STRUCC program. Experiment planning will be approached
from the current PLAN and LOOK commands, which search the environment
of program- or user-specified structural features and give the
experimentalist information on how the structural candidates vary with
respect to those environments. The mass spectrum prediction and
ranking functions will be completed and transferred to BCPL versions
running aS part of the CONGEN program. The concept of prediction and
ranking will be extended to proton and carbon NMR data using, first,
simple additivity rules and, later, refining the predictions based on
more detailed examination of the structural relationships among the
protons and carbons whose signals are being predicted. We hope that
this approach will be effective in choosing a small subset of highly
plausible structures based on agreement between predicted and observed

spectra, just as the mass spectrum analysis functions have proven
useful.

129
We plan no further development of the GC/MS/Computer system, or
the REACT and MAXSUB programs. Rather, these will be used in
applications to current structural problems in conjunction with
compounds or data from other sources. Incremental improvements will be
Made as necessary if a particular new application demands it.

130
Appendix I. CONGEN Workshop Attendees, Affiliation, Research
Interests, Comments on Program and Export Status.

1) Dr. Henry Stoklosa, E.I. DuPont de Nemours.
Dr. Stoklosa has been affiliated with a group at DuPont
involved with computer applications to chemical —
problems, including computer~aided organic synthesis.
He will soon be involved in another group which might
also be able to make use of the REACT program in
addition to his more general interest in CONGEN.

2) Dr. G.W.A, Milne, National Institutes of
Health. Dr. Milne is currently in charge of the
National Institutes of Health contribution to the
NIH/EPA Chemical Information System. His interests
included not only evaluation of the utility of the
program but also exploration of ways in which CONGEN
might be interfaced to the Chemical Information System.
This effort is described in more detail in Section 2.3.
Dr. Milne offered both praise and criticism. He
praised the program itself but wants good user-level
documentation and a large number of sample problems for
future workshops.

We are currently exploring with him the interest of
NIH in obtaining a version of the current program for
use on the NIH PDP-18 facility.

3) Dr. William Brugger, International Flavors and
Pragrances. Dr. Brugger represents the key person at
IFF Research responsible for computer applications in
their laboratories. Structure elucidation is a major
activity of this company not only in analysis of
natural and synthetic products but also in assessing
the relationships between chemical structure and toxic
properties affecting human health. Dr. Brugger has
access to both DEC VAX and IBM computers, the former
representing the laboratory computer. The status of a
version of CONGEN for these machines has been. discussed
previously. Meanwhile Dr. Brugger and his colleagues
are evaluating CONGEN at SUMEX via the GUEST access
facility.

4) Dr. Douglas Dorman is head of the NMR
laboratory at -Lilly Research Laboratories and works
closely with mass spectroscopists and other chemists in
solving structures of a variety of compounds related to
existing or new products. Dr. Dorman has been familiar
with the "old" (non-exportable) version of CONGEN and

131
thus was able to critique the new program not only on
its merits but also on comparison with the old version.
His detailed critique is included in Appendix III. It
is worthwhile to point out that we are in the process
of implementing two of the important new features he
mentions because of their general importance. The
SURVEY and EXAMINE commands will allow both user-
specified features used to explore structural
possibilities and Boolean statements used to select
structures with combinations of features. The MSRANK
facility will perform the selection of plausible
structures based on agreement of predicted and observed
spectra. Dr. Dorman has access to a PDP-19 computer
operating under the TOPS-10 operating system. About
two weeks ago he received acopy of the exportable
CONGEN and is now using the progran in his own
laboratory. He has agreed to provide us with
continuing comments and criticism in exchange for
receiving updated versions of the program.

5) Dr. Jon Clardy, Cornell University. Dr. Clardy
is a recognized leader in development and applications
of the technique of X-ray crystallography in structure
elucidation. His attendance of the workshop was based
on an interest in learning about alternative, computer-
based approaches to the problem. As his letter points
out, the ability to use CONGEN to help solve structures
before expenditure of time and effort in X-ray analysis
would be an important benefit. A more important
outcome of the workshop for future research were
discussions on the possibility of coupling CONGEN-
suggested structures to Patterson search techniques.
In principle, each of the candidates suggested by
CONGEN could be used in turn to guide the search of the
electron density maps for a fitting of the structure to
the maps. The correct structure should yield the least
ambiguous fit. Dr. Clardy will access CONGEN at SUMEX
via the GUEST facility. His ability to use CONGEN at
Cornell depends on the success of the efforts of Mr. In
Ki Mun (next section) .

6) Mr. In Ki Mun, Cornell University. Mr. Mun
attended the workshop representing Prof. Fred
McLafferty at Cornell. Prof. McLafferty’s group has
had for many years an interest in use of computer
techniques to help solve structures, based primarily on
Mass spectral data. His research in this area has led
tO programs which suggest the vresence of
functionalities in an unknown molecule. CONGEN can, in

132
principle, complete such a schema for. analysis by
piecing together the inferred functionalities. Mr. Mun
attended to explore the feasibility of this approach
and to learn how best to use CONGEN on the IBM system
at Cornell. He is currently evaluating the effort of
modifying CONGEN for an IBM version of BCPL. If
successful then this version should be available for
other persons who have good interactive services
available at their IBM installations.

7) Dr. Reimar Breuning, Munich. Dr. Breuning
learned of the existence of the workshops from
discussions with Prof. Djerassi at the IUPAC meeting on
natural products. Dr. Breuning had the opportunity to
attend as he was in the process of arranging a post-
dectoral appointment with Prof. Nakanishi (see next
section). He is actively involved in natural products
structure elucidation at Munich and expects these
interests to continue. Dr. Breuning can access CONGEN
via Guest for the near future from Munich. At Columbia
he will have the opportunity to take advantage of
access to the facility there.

8) Dr. David Lynn, Columbia University. Dr. Lynn
attended the workshop representing Prof. Koji
Nakanishi, the latter a recognized expert in the area
of structure elucidation of anumber of classes of
natural products of relevance to human health. Dr.
Lynn is to act as the focal point for introduction of
the computer methods to that research group. Prof.
Nakanishi’s group has access to a DEC-29 system
Operating under the TOPS-29 operating system. We are
Currently arranging for aversion of CONGEN to be
installed there. We anticipate no problems because of
our successful experiment during the workshop of
cunning CONGEN on a DEC~-28 system at Rutgers.

9) Dr. Y¥. Gopichand, University of Oklahoma. Dr.
Gopichand attended the workshop representing the marine
natural products group of Prof. Francis Schmitz. This
group specializes in structure elucidation of
halogenated terpenoid molecules possessing a variety of
biological activities and marine sterols representing
intermediates or end products in steroid biosynthesis.
As evidenced by the letter of critique, this group
represents an axcellent example of classical approaches
to structure elucidation; sufficient data are collected
such that the number of structural possibilities is
reduced to avery small number. As the letter also

133
indicates, CONGEN should be useful at least to check
the rigor of their structural assignment. What should
prove more interesting is whether or not such a group,
with little computer expertise, can use CONGEN earlier
on in the process of structure elucidation to guide
subsequent collection of data.

18) Ms. Wendy Harrison, University of Hawaii. Ms.
Harrison attended the workshop representing the marine
natural products group of Prof. Paul Scheuer at Hawaii.
This group is engaged in structure elucidation problems
similar to those encountered in Prof. Schmitz‘s
laboratory, although focus is on different classes of
organisms. Their letter of critique mentioned that due
to absence of critical high resolution mass
spectrometric data to establish molecular formulas,
they have been unable to use CONGEN this past month to
help on their problems. This situation is expected to
be resolved soon. Prof. Scheuer plans to use CONGEN
as an aid in a course in structure elucidation this
coming semester. One difficulty is that this group
only has access to a DEC PDP-1ll system. This
introduces all of the problems mentioned earlier ona
version of the program for mini-computers, including
proliferation of manufacturers and operating systems.
For example, there is a version of BCPL for PDP-ll
series machines, but only under the RI-11 operating
system, whereas the Hawaii group runs under The RSX-11
Operating system. For the near future, access from
Hawaii will be to CONGEN at SUMEX running umder the
GUEST directory.

11) Dr. Laszlo Tokes and Dr. Michael Maddox,
Syntex Research. Drs. Tokes and Maddox are,
respectively, in charge of the mass spectrometry and
NMR laboratories at  Syntex Research. They are
responsible for the majority of structure elucidation
problems which rely on physical methods. Their
interest in CONGEN is that it might help them solve
certain problems in less time than required by manual
methods. Syntex research has access only to laboratory
mini-computers such as the DEC PDP~1l series machines.
They, too, must await a smaller version of CONGEN
suitable for mini-computers before being able to use
the program in their own laboratory.

12) Dr. John Figueras, Kodak Research Laboratory.

Dr. Figueras attended representing the Analytical
Sciences Division of Kodak’s Research Laboratory. This

134
division is responsible for data collection and
analysis in support of the structure elucidation
activities of the Laboratory including not only new
developments in the photographic process but also the
new technology of thin-film bound enzymes systems for
clinical analyses. His role was to evaluate what part
CONGEN could play in the on-going automation of of the
Division. The Division will be obtaining a DEC-28
system in the near future on which CONGEN will run
directly. Meanwhile we have offered GUEST access in
return for continuing critique on utility of CONGEN for
large, poly-heteroatomic, aromatic molecules of the
types encountered in their research.

13) Dr. Charles Snelling, University of Illinois.
Dr. Snelling attended the workshop representing Prof.
Kenneth Rinehart in the Chemistry Dept. at Illinois.
Prof. Rinehart is also an acknowledged expert in
structure elucidation with emphasis on macrolide
antibiotics, halogenated terpenoids and other classes
of natural and synthetic products of relevance to human
health problems. Dr. Snelling had several comments and
criticisms about the difficulties facing a novice user
of such a‘complicated system. Although he found the
program useful and wishes to get it running at
Illinois, he would like to see some changes made to
simplify use of the program for the chemist. There are
many computer systems available at Illinois and choice
of which system on which to mount CONGEN will depend on
the ease of the task and access to a particular system.
This is currently under study by them.

14) Dr. Gilles Moreau, Roussel UCLAF. Dr. Moreau
attended the workshop representing the French
pharmaceutical concern Roussel UCLAF. This company
Maintains an active group in computer applications in
chemistry and wished to evaluate CONGEN for its use in
their structural problems. The company is quite
interested and will explore use of the program via
GUEST access. They are concerned about issues of
secrecy for new problems and we have made it quite
clear that GUEST access represents public knowledge of
their problems. Therefore, assuming their interest
continues, we will be arranging some alternative to
SUMEX for use of CONGEN. They have access to IBM
equipment and we are awaiting further description of
the form of access possible.

15) Prof. Andre Dreiding, Zurich. Dr. Dreiding

135
has been interested in both the problem-solving and the
pedagogical aspects of CONGEN for some time. He had
previously used the old version and was gratified to
see the improvements in the new version. He would like
to see much more attention paid to the actual
structures (i.e., in three dimensions) of molecules
rather than simply their constitutions as CONGEN, with
the exception of the STEREO command, currently
represents structures. We are, of course, working very
hard to introduce concepts of stereochemistry into our
computational procedures. Prof. Dreiding will probably
be able to access CONGEN at Zurich on an existing PDP-
1@ installation. If so, export to his group will be
trivial.

16) Dr. James Shoolery and Dr. Michael Gross,
Varian Associates. Dr. Shoolery is in charge of
Varian’s NMR application laboratory and Dr. Gross is in
charge of computer software for Varian’s NMR/computer
systems. Their respective interests match their
responsibilities. Dr. Shoolery feels that CONGEN could
be avaluable assistant in helping solve structures
based primarily on proton and carbon NMR data. In
fact, he has been able to demonstrate such an approach
on some recent problems of persons outside Varian. Dr.
Gross is interested in a mini-computer version of
CONGEN for incorporation into their existing data
system. Although we cannot ourselves support such an
effort, we have agreed to provide them with program
listings and documentation to explore the feasibility
of a small machine version.

17) Dr. Daniel F. Chodosh, Smith, Kline and
French. Dr. Chodosh was not actually invited to the
‘workshop, but happened to visit our group during one of
the sessions. He was sufficiently impressed that he
procured a tape to carry away a copy of the program
with him. He has now been supplied with a version of
CONGEN for the PDP-1@ and as of last week has it
running at the SKF research laboratories in
Philadelphia. This is an interesting experiment
because as an informal attendee of a small portion of
the workshop, he is learning the program almost from
scratch based on existing help facilities and
documentation. .He will begin introducing others to the
program when he has developed sufficient familiarity to
be comfortable with the program.

In addition, the following persons during the past year have asked

136
for information about and access to CONGEN. For the most part we have
granted access through the GUEST directory, setting up an account only
for those users with more than occasional log-ins.

Dr. David Cowburn
Physical Biochemistry

The Rockefeller University
New York City

We have sent Dr. Cowourn information on access to CONGEN and are.
currently discussing how to use some of our computational methods
or extensions to them for assistance in his peopiems of elucidating
peptide conformations.

Douglas Henry

School of Pharmacy
Oregon State University
Corvallis, Oregon

He has been sent our programs for structure drawing for use
on his own computer.

The following have asked for and received information on access to
CONGEN at SUMEX via the GUEST facility.

Dr. H. Kating

Institut fur Pharmazeutische Biologie
Der Universitat

Bonn, Germany

Dr. Kerber
Lehrstuhl D fur Mathematik
Aachen, Germany

Or. Brenda J. Kimble
Radiobiology Laboratory
University of California
Davis, California

Dr. J. Neubuser
Lehrstuhl D fur Mathematic
Aachen, Germany

Dr. George Padilla

Dept. of Physiology

Duke University: Medical Center
Durham, N.C,

Dr. W. Sieber
Sandoz Ltd.

137
Basel, Switzerland

Dr. Babu Venkataraghavan
Lederle Laboratories
Pearl River, New York

We also helped him bring up the Fortran draw program on the DEC~19
system at Lederle

Dr. Stephen Wilson
Dept. of Chemistry
Indiana University
Bloomington, Indiana

138
Appendix II. Sample Letter of Invitation Sent

To Prospective Workshop Attendees.

139
LoPY

STANFORD UNIVERSITY
STANFORD, CALIFORNIA 94305

DEPARTMENT OF CHEMISTRY

August 24, 1978

Professor Kenneth Rinehart
Department of Chemical Science
University of Illinois

Urbana, Illinois 61801

Dear Professor Rinehart:

I am writing to determine your interest in a mnini-workshop we plan to
hold at Stanford on use of computer programs for computer~assisted
structure elucidation. Over the past three years we have been involved
in a research effort directed in part to exploring the feasibility and
utility of interactive computer programs as tools to help chemists
solve unknown structures. The most highly developed of these research
tools is the CONGEN program, and I know that you have in one way or
another been exposed to this program.

We have made CONGEN available on the SUMEX computer here at Stanford
via communications networks to a selected group of chemists because we
know that the only way to improve the program and to make it useful is
to apply it to a variety of real structural problems in chemistry. We
have learned a great deal from this experience and have designed a
production version of CONGEN with the primary goal of eliminating the
deficiencies in the research version. In particular the new program is.
much simpler to use, much gpaller and faster and can be exported to
certain other computers.

There is a significant amount of work which remains to be done to make
the exportable version of CONGEN a truly useful chemist's "assistant".
Some of this work is underway now. But before investing a great deal
of time and effort in polishing a new version to make it more useful
and acceptable, we would like to expose a group of several chemists
experienced in structure elucidation to this version of the program and
base future research and development efforts on their perceptions of
deficiencies remaining in the program and its ultimate utility. We
feel its usefulness can be significant in terms of exploring
alternative structural possibilities and in guaranteeing no plausible
alternative has been overlooked. However, in order to demonstrate that
utility we need assistance in developing a version which large numbers
of chemists can use easily and productively in their own laboratories
and on their own computers. We feel that by participating in this
workshop you can contribute in an important way.

We plan an approximately four to five day informal workshop here at
Stanford, to be attended by you or one of your experienced co-workers.
During this time you would be able to use the exportable version of
CONGEN here at SUMEX to work on recent structural problems encountered
in your laboratory, which we hope will include some which would be

140
unknowns. In turn, we would be learning how to finish development of
the first version of the program for export to your laboratory with
some guarantee that major problems or deficiencies would be eliminated
based on the experience of the workshop. We have not yet established a
time for the workshop, but we are tentatively considering a date
between October and December, 1978. We are also trying to arrange for
travel support for non-industrial persons. We are seeking now only an
expression of interest on your part.

There are only two requirements on your part, other than your expressed
interest in participating. The first requirement is that you must be
capable of accessing CONGEN at SUMEX or preferably be capable of
running CONGEN on your own computer system. The second requirement is
that you actually be interested in using the program in the future
(assuming the workshop was worthwhile) to provide us with some feedback
on whether improvements meet your criteria for acceptance.

Currently the new version of CONGEN can be used on Digital Equipment
Corp. PDP-10's and 20's with little problem. We are now exploring use
of IBM 360 and 370 series machines. We have already demonstrated that
small segments of the program run on suitably configured IBM computers.
However, for CONGEN to be used interactively requires a time-sharing
operating system. If you have access to an IBM computer but are
uncertain about the operating system, find out from your computer
systems people the name of the operating system and call or write te us
with that information.

While at Stanford you would also be able to take a close leck at some
of our other research efforts which are not quite so far along as
CONGEN , including a recently finished stereochemical structure
generator, the REACT program anda variety of tools for ranking
structures based on comparison of predicted and observed spectra,
facile examination of large sets of structural candidates for common
and unique features and methods for assistance in experiment planning.
We would appreciate your evaluation of these efforts also. The main
emphasis, however, will be on CONGEN.

Obviously we do not expect a firm commitment for attendance at this
time, but if at all possible I would appreciate a reply. Indications
of your interest are important in this regard.

Yours sincerely,

Carl Djerassi
Professor of Chemistry

141