The main objective of the project is the development of a protocol capable of identifying i) novel drug binding sites and ii) novel drug-protein complexes of Mycobacterium tuberculosis proteins. Our present knowledge of human genomics and proteomics a llows possible new approaches in drug discovery, motivated by combinatorics-based data-mining and knowledge discovery techniques, sometimes labelled as "high throughput" techniques. Solely these approaches (without sound biological background) seem to be still inadequate for drug discovery. In the present project we aim to develop a protocol, rather than a software with two ingredients: a, A method, with an algorithmic solution in its centre, which identifies surface indentation patterns in protein 3 D structures. Using very high throughput methods based on recent result of data mining and information retrieval, pairwise searches will be conducted to predicate drug-protein binding in two main phases, a rough first phase (Silico 1) and a more thoroug h docking investigation (Silico 2) in the second phase. b, A structural and molecular biology protocol for examining the most promising drug-protein fit pairs with a variety of medium throughput and low throughput methods.