Synthetic accessibility estimation

Gillet, V.J., Myatt, G., Zsoldos, Z., and Johnson, A.P. SPROUT, HIPPO, and CAESA tools for de novo stmcture generation and estimation of synthetic accessibility. Perspect. Drug Disc. Des. 1995, 3, 34-50. 

Ertl, P. and Schuffenhauer, A. (2009) Estimation of synthetic accessibility score of drug-like molecules based on molecular complexity and fragment contributions. Journal of Cheminformatics, 1, 8. 

Virtual chemistry is taken to mean the universe of compounds we could make but haven t. An example would be a combinatorial chemistry library, from which only a subset is synthesised. In principle, any other member of the library is also synthetically accessible. The size of the universe is the product of the numbers of R-groups and can quickly reach infeasibly large numbers. Workers at Pfizer estimate that this accessible universe just within Pfizer is 10 compounds, compared to the < 5 million compounds in the archive. It is impractical to explicitly enumerate all these compounds and to search them, so two methods have been developed to narrow down the search space. Lessel et describe further improvements to the colibri... 

V. J. Gillet, G. Myatt, Z. Zsoldos, and A. P. Johnson, in Perspectives in Drug Discovery and Design, K. Muller, Ed., ESCOM, Uiden, 1995, Vol. 3, pp. 34-50. SPROUT, HIPPO, and CAESA Tools for De Novo Structure Generation and Estimation of Synthetic Accessibility. 

SPROUT, HIPPO and CAES A Tools for De Novo Structure Generation and Estimation of Synthetic Accessibility. 

Estimation of Synthetic Accessibility Score of Drug-Like Molecules 291 Based on Molecular Complexity and Fragment Contributions... 

A method to estimate ease of synthesis (synthetic accessibility) of drug-like molecules is needed in many areas of the drug discovery process. The development and validation of such a method that is able to characterize molecule synthetic accessibility as a score between 1 (easy to make) and 10 (very difficult to make) is described in this article. 

The major problem when developing methods for estimation of SA is the validation of results. It is not straightforward to extract synthetic complexity out of the protocol describing molecule synthesis. While the overall yield over the sequence of synthetic steps gives some information, this depends also on the effort which has been undertaken to optimize the process and if only low amounts are needed for initial experiments, then a non-optimal synthesis is tolerable. Another possible measure of synthetic accessibility of a molecule could be its price in catalogues of chemical providers. The price, however, depends on too many factors not related to SA (for example novelty of the reagent, demand, packaging. 

Despite the good performance of the SAscore documented above, we are well aware also of limitations of this method. The SAscore cannot compete with more sophisticated approaches for estimation of synthetic accessibility which reconstruct the full synthetic path, in cases when the results are critical, for example... 

Gillet VJ, Myatt G, Zsoldos Z, Johnson AP SPROUT, HIPPO and CAESA Tools for de novo structure generation and estimation of synthetic accessibility. 

CAESA = Computer Assisted Estimation of Synthetic Accessibility CSD = Cambridge Structural Database DLD = dynamic ligand design FEP = free energy perturbation MCSS = maximum common substructure PDB = Protein Data Bank. 

It is important to stress that an experienced synthetic chemist could perform as well as CAESA in estimating the synthetic accessibility of a small set of molecules. However, the CAESA system has the advantage when it comes to ranking hundreds or even thousands of molecules. 

Chemical space—that is, aU possible molecules—is estimated to be greater than 10 ° molecules with 30 or fewer heavy atoms 10 pg of each would exceed die mass of the observable universe. This figure decreases if criteria for synthetic accessibility and drug likeness are taken into account and increases steeply if up to 35 heavy atoms (about 500 Da) are allowed. Positing even a modest specificity of proteins for dieir ligand, the odds of a hit in a random selection of 10 molecules from diis space seem negligible. 

---