Mini Reviews in Medicinal Chemistry - Volume 4, Issue 10, 2004

Decision support systems have been widely used in drug discovery owing to the complexity of data and information involved. In this article, we review both commercially available packages and publicized in-house made systems. All selected systems should be able to handle chemical structures and to organize information for decision makers, more specifically for medicinal chemists. Although we do not rank these system, pros and cons of these systems will be discussed.

Molecular similarity and diversity analysis has played a significant role in computer-aided drug discovery for more than a decade. Compound classification methods have also become increasingly important for the design and organization of compound databases and in silico screening. Here we review these related methodologies and discuss selected applications.

The contribution of physicochemical descriptors to lipophilicity, water solubility, and intestinal absorption and oral bioavailability in humans is considered. Partitioning in the octanol / water system is presented as a competition between two opposing effects: volume and hydrogen bond acceptor ability. Water solubilities of liquid compounds are roughly equal to their reciprocal logP values. However, there is also a detectable contribution of H-bond donor ability to water solubility. The main problem in predicting the solubilities of solid chemicals and drugs is the estimation of their crystal lattice energies. QSAR approaches that add terms such as melting point, and the product of H-bond donor and acceptor parameters are not sufficient to make these predictions practical. Human intestinal absorption for passively transported drugs is almost completely correlated with hydration processes that are determined by H-bond acceptor and donor abilities. It is emphasized that structural features of drug molecules have significant influences on their properties. Classic QSAR approaches are not enough to create stable, predictive models for diverse drugs. A combination of Similarity and QSAR approaches is one possibility to take all physicochemical properties in addition to structural features into account.

Advances in high-throughput virtual screening using docking, predictive ADME methods and their integration with informatics and high-performance computing are reviewed. Docking approaches have led to the identification of novel active compounds. Predictive ADME methods have improved on selective test sets with broader training sets, though extensive validation is lacking.

The introduction of combinatorial chemistry groups into pharmaceutical companies provoked a desire for efficient and effective methods for library design and optimisation. This, in turn, has resulted in a large number of scientific publications, detailing a variety of approaches to the problem. This review attempts to describe the major works in the literature, to set them in context both chronologically and scientifically, and to identify the outstanding challenges that must be addressed, if this area of research is to maintain the rapid progress seen hitherto.

Microtubules (MTs) play important and diverse roles in eukaryotic cells. Their function and biophysical properties have made α-and β-tubulin, the main components of MTs, the subject of intense study. Interfering with normal MT dynamics, for example, by the addition of tubulin ligands, can cause the cell great distress and affect MT stability and functions, including mitosis, cell motion and intracellular organelle transport. It has been shown in the literature that tubulin is an important target molecule for developing anticancer drugs. Tubulin binding molecules have generated considerable interest after the successful introduction of the taxanes into clinical oncology and the widespread use of the vinca alkaloids vincristine and vinblastine. These compounds inhibit cell mitosis by binding to the protein tubulin in the mitotic spindle and preventing polymerization into the MTs. This mode of action is also shared with other natural agents eg colchicine and podophyllotoxin. However various tubulin isotypes have shown resistance to taxanes and other MT agents. Therefore, there is a strong need to design and develop new natural analogs as antimitotic agents to interact with tubulin at sites different from those of vinca alkaloids and taxanes. This minireview provides SAR on several classes of antimitotic agents reported in the literature. The structures and data given are essential to the scientists who are involved in drug design and development in the field of anticancer drugs.

New drugs are introduced to the market every year and each individual drug represents a privileged structure for its biological target. In addition, these new chemical entities (NCEs) not only provide insights into molecular recognition, but also serve as drug-like leads for designing new future drugs. To these ends, this review covers the syntheses of 23 NCEs marketed in 2003.