SOS (Simulated Organic Synthesis) is a “classical” computer-aided organic program, working step by step from the target to the starting materials (backward approach) generating a retrosyntetic tree. It was first developed for heterocyclic chemistry. It was then improved by coding reactions by means of mechanisms for all fields of organic synthesis.

How does it works ?
The program searches for a reaction, i.e. a substructure, and if it is present the corresponding precursor is built and displayed on the screen. There are three modules in SOS : a drawing one to input the target, another to draw the reactions and SOS itself which generates the precursors.
Example of the Diels Alder reaction :

The substructure to be found is the cyclohexene ring. If it is present in the target SOS replaces it by the precursor substructure, here diene and dienophile. For example, in the structure below SOS perceives the cyclohexene and the following solution is proposed :

Some problems, however, can arise. For example, if we propose this molecule to SOS, the solution will be :

which is impossible due to the presence of the allenic function in the five-membered ring. So, another step is necessary to code a reaction : an evaluation module which discards erroneous solutions. This evaluation was performed by means of simple tests such as :
If (atom/bond) is (substructure) then (discard/goto another test/end) else (discard/goto another test/end).
These tests are unable to discard all the bad solutions and the program can make mistakes. Our objective, however, was not to find the best solutions, but rather to propose original ideas to the chemist..
Actually a reaction is coded in three parts :
1) Search of a substructure caracteristic of the reaction.
2) Instructions to built the precursor(s).
3) An evaluation module to eliminate simple non viable solutions.

It is possible to generalize the reactions by using the symbol A for any atom, or T for heteroatom, it is also possible to indicate that a bond could be simple or double, or double or triple, etc. For example, for the Cope reaction it is possible to generalize this reaction : all atoms are set as A, bonds 1 and 5 can be double or triple, and bonds 2 and 4 can be simple or double.
The reaction is coded by the following scheme :

When it starts SOS analyzes the target and searches for important features such as rings, nucleophilic centers, etc. So it is possible to code mechanisms such as nucleophilic substitution or nucleophilic addition with one scheme :

L stands for any leaving group. Thereby SOS proposes a general plan and not a precise solution, but interpretation of which does not put problem to the chemist. This approach allows the program to propose new solutions even if they have not been described in the litterature.

The results of such coding can be seen with the solutions given below for the thiazole ring.

Some original suggestions which show the “creativity” of SOS :

SOS can be summarised by the following simplified flowchart :

The very first version of SOS was written in Fortran for a 1130 IBM computer (16 Kwords of memory).
Later it was adapted in BASIC for an Apple II microcomputer, and finally a QuickBasic version was developed for IBM/PC computer. ( 1994 : A version of SOS with stereochemistry for IBM/PC was developed (André Bertrand PHD)).
A Windows version was under development.

Papers about SOS :

1 -    Emploi de l'ordinateur en synthèse organique.
         R. Barone
         Chimie Actualités, 1973, 1515, 29-32.
2 -    Assistance de l'ordinateur en synthèse organique. Réactions d'hetérocyclisation.
         R. Barone, M. Chanon et J. Metzger
         Revue de l'I.F.P., 1973, 28, 771-790.
4 -     Ordinateur et synthèse organique.Utilisation des mécanismes réactionnels.
         R. Barone, M. Chanon and J. Metzger
         Tetrahedron Letters, 1974, 32, 2761-2764.
5 -    Ordinateur et synthèse organique. Application d'un programme non interactif à la synthèse du thiazole.
         R. Barone, M. Chanon and J. Metzger
         Chimia, 1978, 32(6), 216-219.
6 -     Ordinateur et synthèse organique. Représentation des molécules et des réactions. Propositions de synthèse pour l'aza-6-uracile.
          R. Barone and M. Chanon
          Nouveau Journal de Chimie, 1978, 2(6), 659-663.
8 -     Synthèse organique assistée par ordinateur. Applications aux indazoles.
         R. Barone, P. Camps and J. Elguero
         Anales de Quimica, 1979,75, 736-738.
15 -   Microcomputers and organic synthesis.
         R. Barone, M. Chanon, P. Cadiot and J.M. Cense
         Bull. Soc. Chim. Belg., 1982, 91(4), 333-336.
20 -   Microcomputers and organic synthesis. Application of an interactive program to the photochemical synthesis of pheromones.
         R. Barone, M. Chanon and M.L. Contreras
         Nouveau Journal de Chimie, 1984, 8(5), 311-315.
23 -   Computer-assisted synthesis : an undergraduate student experiment in organic chemistry.
         M.P. Bertrand, H. Monti and R. Barone
         Journal of Chemical Education 1986, 63, 624.

[Home] [Some dates] [Papers] [sos] [marseil] [conan] [holowin] [graal] [miscellaneous]