Drug Discovery

Abstract

Despite availability of several drugs, a vaccine, decades of research and concerted efforts from medical and allied communities to manage tuberculosis, it is clear that the Mycobacterium tuberculosis has been successful in defying these efforts and continues to pose a major threat to mankind. Newer approaches, in particular, newer strategies for drug discovery are therefore urgently required. The science of drug discovery has witnessed multiple paradigm shifts in the past few decades, from a predominantly ligand-centric approach to a target-centric approach and now recently leaning towards a systems-based approach. The shifts can be attributed to several factors such as availability of publicly accessible databases containing genome sequences, functional and structural data of macromolecules, high-throughput experimental profiling, protein–protein interactions and pathway models, as well as adaptation and application of computational methods for efficient data mining and modeling. Several omics-scale experimental and in silico approaches have emerged recently to systematically address important questions in biology, with an obvious impact on drug discovery.

A systems view enables a broad understanding of the system as a whole, providing significant insights at multiple stages in the drug discovery pipeline, from target identification, understanding pharmacokinetics and pharmacodynamics, to personalized medicine. Of the systems approaches for drug discovery, modeling metabolism in the causative agent has received some attention. Flux balance analysis, and metabolic control analysis that can simulate the relative reaction fluxes under a variety of conditions, have provided lists of predicted essential proteins and hence potential drug targets. Perturbations such as gene knock-outs, drug inhibitions, double and triple knock-outs, exposure to different chemical environments can all be modeled through this approach. Interactomes capturing structural and functional protein–protein linkages have been useful in identifying proteins strategically located in the network, which when inhibited would perturb the network significantly. There have also been examples of rule-based or logic-based modeling studies that will help in identifying the effect of different scenarios of host–pathogen interactions and adaptations within each, thereby identifying optimal strategies for therapeutic intervention. The models themselves are increasingly being enriched with experimental information, as more and more genomics and proteomics data is becoming available. The potential of these methods that still remains to be tapped in drug discovery programs are discussed. The stage seems set for the integration and application of skills from mathematics, computer science, and engineering disciplines, to address complex problems in biology and drug discovery, in a big way.