be: Clair Cheer

STANFORD UNIVERSITY
STANFORD, CALIFORNIA 94305

DEPARTMENT OF CHEMISTRY

July 16, 1976

Po. A. G. Gonzdélez

Departamento de Quimica Organica
Universidad de La Laguna

Instituto de Investigaciones Quimicas
C.S.1.C.

Tenerife, Spuin
-, f
Dear Prof. Gonzdlez:

| have been following your work in the natural products field for quite some time and
have been impressed by the variety of interesting compounds you have worked on and
the quality of the work. Just as we have done in our own work, you have been very
dependent on a variety of spectral methods, and | thoughi that you might be interested
in a computer approach that we have called CONGEN and which we have been utiliz-
ing with some success in our own structure studies. This work has been supported by
grents from the NIH with the purpose of making this program and general approach more
available to the chemical community in this country and abroad. | thought that you
might be sufficiently interested to be tempted to use it yourself, and it is for this
reason that | am writing to you. Our initial work on CONGEN has been published in
J. Amer. Chem. Soc. 97, 5755 (1975), and if you have not read it earlier | would
suggest that you look at it in order to give you background for what follows in this letter.

| thought that you would be particularly interested if we applied the CONGEN approach
to one of your problems in order to illustrate its power, and for that purpose Dr. Jean-
Claude Braekman (who works at the University of Brussels but who came here for a few
weeks to learn how to handle CONGEN himself) selected a couple of problems outlined
by you in two recent papers. The first one deals with your paper on the constituents

of Laurencia perforata, published in Tetrchedron Letters 2499 (1975). You will note from
the attached material that in addition to the structure A proposed by you for perforatone
there are seven altematives that seem to fit your data and are also based on the skele-
ton Hl. There are also a dozen other structures based on skeletons | and {1 that fit,

and | am curious whether you considered these.

In the case of perforenone A and B there appear to be only two structures other than
the one that you selected, and while the one with the cyclobutene ring seems unlikely,
for the sake of rigorousness one should probably try to find some evidence to exclude
it unambiguously.
Prof. A. G. Gonzdlez
July 16, 1976
page 2

Dr. Braekman has also applied the CONGEN program to furocaespitane, which you
published in Tetrahedron Letters 3625 (1973), and here the program generated five
additional structures over and above the one proposed by you which seemed to be con-~-
sistent with your data.

| would like to encourage you to go over these structures, since there are not very
many of them, to determine whether you have any evidence that would exclude them.
It is quite possible that you may have additional spectroscopic or chemical data which
would introduce additional constraints that would eliminate several, or possibly all but
one, of the structures listed.

The chief reason for Dr. Braekman having gone through your recent Tetrahedron Letters
papers is that we wanted to demonstrate to you the utility of this method and encourage
you to feel free to use our program. It is probably unlikely that you have a computer
terminal in Tenerife with which you can communicate with our main computer here at
Stonford although there are several such terminals in certain cities in Europe whereby
such communication over long distance is easily feasible. However, as we have already
done with several other investigators such as Professor Schmitz from Oklahoma, Professor
_ Minale from Naples, and Professor Nakanishi from Columbia, it is sufficient if you send
us the relevant spectral and chemical data, similar to what you have published in
Tetrahedron Letters, and we would be happy to send you the appropriate computer print-
out. We are doing this because we are anxious to demonstrate additional applications
of this program and furthermore to refine it as we expose different users to it. The
attractiveness of CONGEN is that it is a dynamic program which can be utilized at
various stages of a chemist's work so as to reduce the various structural possibilities

for a given compound. We have ourselves found that frequently by utilizing it at an
early stage structures have been called to our attention which we would have otherwise
overlooked and which could easily be eliminated by some relatively simple additional
spectroscopic or chemical work.

Please do not hesitate to write if you have any questions and feel free to utilize our
computer facilities for your structure work,

Yours sincerely,

Car] Djerassi

Professor of Chemistry
enclosures

ce: Dr. Jean-Claude Braekman