The science of drug discovery has been witnessing multiple paradigm shifts in the recent past due to several factors such as availability of genome sequences, significant growth of databases in molecular detail due to ‘omics’-scale experiments, development of systems-level models as well as adaptation and application of computational methods to biological problems. Systems level understanding has the potential to address several important issues that arise in drug discovery, such as the choice of an optimal target, causes for failure of existing drugs including adverse effects.
Target identification is a critical step in modern drug discovery. Identifying the right target however is by no means simple, since a variety of factors need to be considered simultaneously. One of the main unanswered problems with most drugs is that many of them exhibit adverse drug effects due to additional interactions with unintended host proteins. A systems perspective of the proteome in terms of the interaction profile is essential to understand this aspect. An ideal target for an anti-infective drug should first be essential to the pathogen, and preferably also unique, but should not share similarity in its ability to bind drug-like molecules with proteins from the host. The host and the pathogen genomes can be compared computationally at various levels of abstractions, such as through their gene or protein sequences, protein structures, biochemical function(s) and systems level interactions. Recently, new methods have been developed in the laboratory which enable comparing the host and pathogen pocketomes, through which a measure of targetability can be computed. This presents a rational approach to prioritize targets, explore poly-pharmacological targets, understand drug pharmacodynamics, obtain lead clues and obtain shortlists for exploring drug repurposing. A case study with tuberculosis proteome will be presented, illustrating how such knowledge can be used in the discovery of new, safer drugs.