Library Quality Metrics

In this chapter, the focus is on chemistry. This focus is not confined to ways to make compounds in the screening libraries it encompasses broader issues, such as the quality of the bbrary, its biological relevance, and such andal practical matters as compound purity, stability, and solubility. The pivotal role of the compound library in every pharmaceutical company is illustrated in Fig. 28.1, where the drug discovery and development process is flanked by the two key disciplines, biology and chanistry. 

Drug Discovery and Development, Volume 2 Drug Development, Edited by Mukund S. Chorghade Copyright 2007 John Wiley Sons, Inc. 

At this moment the attrition rate in clinical trials is far too high. If, better-quality candidates are developed in preclinical research, the likelihood that they will pass the clinical trials will be increased. The cost savings in such a, case will be enormous. A recent study has revealed that more than 40% of the candidates in clinical trials fail due to poor (bio)pharmaceutical properties and more than 20% due to toxicity. These reasons depend heavily on the chemistry of the candidate and probably could be identified in preclinical research. The rate-limiting aspect of drug discovery is not in finding the lead it is turning the 

---

In this chapter we focused on metrics designed to improve the quality of a compound library. Each metric can be rationalized most can be measured. Serendipity is not necessarily... 

A second measure of the quality of simulated speara can be determined by performing a library search. In the search procedure, each simulated spectrum was compared to all speara contained in the spearal library to determine the most similar measured spectrum. The spectral library, compiled from data in the literature, contained 521 spectra from several chemical classes, including norbornanes and norbornanols, cyclic and acyclic alkanes and alcohols, cyclic ketones, steroids, PCBs, and an assortment of small molecules. The squared Euclidean distance metric was used to assess similarity. Sorted chemical shift values of each simulated and library speartun were compared, and the five library spectra having the smallest squared Euclidean distances for each simulated spearum were recorded. For each of the 32 reference compounds, the library spectrum that was retrieved with the smallest squared Euclidean distance, when compared with the simulated spearum, was its corresponding observed spectrum. Thus, in the library search, the correa spectrum was always retrieved as the most similar spearum to the simulated spectrum. 

---