Structure-assisted lead optimization

Many predictive, validated models have been developed using these QSAR techniques and have often assisted in the selection of structures for lead optimization. Often the QSAR results are not available until after the process of lead optimization has already progressed, and these models represent retrospective analysis of the lead optimization process rather than a direct influence on the design of the lead optimization compounds. 

The second question is of prime importance in lead optimization, and structural information can greatly assist this process. X-ray crystallography and NMR can equally and in... 

The simplicity of Hansch analysis also means that experienced medicinal chemists may be able to identify trends in activity without the assistance of a QSAR equation. Making individual new lead analogues is generally a slow process, and a medicinal chemist has ample time to examine SAR data. While a chemist will not be able to quantify a structure-activity relationship, just knowing the approximate trend of the relationship is usually adequate for lead optimization. Hansch analysis is valuable only if it can reveal something that is not already known about the compounds being tested. 

Han van de Waterbeemd, Bernard Testa, and Gerd Folkers, Computer-Assisted Lead Finding and Optimization. Current Tools for Medicinal Chemistry, in European Symposium on Quantitative Structure-Activity Relationships, Vol. 11, Wiley-VCH, Weinheim, 1997. 

Johnson, S. R. and Jurs, P. C. (1997) Prediction of acute mammalian toxicity from molecular structure for a diverse set of substituted anilines using regression analysis and computational neural networks. Comput. Assist. Lead Find. Optim., [Eur. Symp. Quant. Structure-Activity Relationships QSAR and Molecular Modeling], 11th, pp. 31 8, Lausanne, Switzerland. 

H. Kubinyi, in Computer-Assisted Lead Finding and Optimization Proc. 11th European Symp. on Quantitative Structure-Activity Relationships, Lausanne, 1996, eds. van de H. Waterbeemd, B. Testa, and G. Folkers, Verlag Helvetica Chimica Acta and VCH Basel, Weinheim, 1997, pp. 7-28. 

The application of QSAR to bioactive synthesis has always suffered from an unfortunate paradox. In order to develop a useful equation, it is necessary to first complete a substantial fraction of the synthesis. Only then can the derived equation assist in extending or optimizing the bioactive series. No help is available for the earliest or intermediate stages of synthesis which have already been passed. Nor is it certain that a useful equation can be gained from the first 10-20 members of a series. Poor selection of structural changes, variable biodata, differential metabolism of some members and the presence of unknown factors can all lead to poor correlations of little practical use. These problems are common to anyone who has attempted QSAR on novel bioactive series. 

It often happens during chemical modification and optimization that compounds exhibit unexpected unfavorable effects, even when the lead compounds are free from such effects. In such cases the most practical way to reduce the unfavorable effects is further chemical modification. To perform effective and efficient chemical modification, it is important for medicinal chemists to understand the structural factors responsible for unfavorable effects. To practically assist medicinal chemists in understanding these structural factors and to guide their efforts to reduce unfavorable effects through chemical modification, predictive models should be simple and easy to understand. 

The processes as just described do not require any information coordination between the genomics and proteomics work and the subsequent lead discovery and optimization. In the last decade, there has been a growing effort to transition the information from the proteomics step into a computer-aided dmg discovery process—to use the information about the protein itself to help choose appropriate compounds to screen for the desired efficacy. The use of computers to assist in drug discovery neither is new nor was always tied to the use of protein structures. The recent developments in proteomics have focused the work on protein structure-based computer-aided drug design. 

Another field for assisting in the selection of potential lead compounds or optimized leads is called Quantitative Structure Activity Relationships (QSARs). These methods determine correlations between certain descriptors of the molecular structures and the measured biological activity to produce a predictive model. That model can then be used to predict the activity of other structures prior to actually testing. 

Mass screening of our compound collection, utilizing the Amersham SPA technology, afforded pyrone and coumarin non-peptide templates as initial lead structures. X-ray cocrystallization and structure-based design were utilized to assist in the design of more potent inhibitors. These efforts resulted in the design of the 5,6-dihydropyrones, which afforded a more flexible template from which to fill the internal pockets of the enzyme. Optimization of the dihydropyrone series afforded a potent antiviral agent,... 

---