Applications of quantitative pharmacodynamic effect markers in drug target identification and therapy development

Genome-wide transcriptional profiling is now feasible, and profiling of the proteome, although technically challenging, is advancing rapidly. Expression profiling provides a tool to accelerate discovery in a broad range of sciences, but its greatest impact on human health may be on the process of drug discovery and therapy development, and investigation of the functional networks underlying drug responses of diseased and normal tissue. For anticancer agents in particular, antitumor effects and toxicities to critical normal tissues may rest in a delicate balance that is governed by complex pharmacokinetic (PK) and pharmacodynamic (PD) inter-relationships. Recent advances in the development of mechanistic computational PD models promise to promote an understanding of these interrelationships, provided suitable quantitative PD effect markers will be identified. Here we describe both advances toward the unsupervised application of PD models to complex expression profiling datasets, as well as approaches to address the technical requirement of these models for quantitative assessment of protein expression levels. Together, these models and analytical approaches may contribute to the rational design of more effective pharmacotherapies.