Figure 1

The targetTB Target Identification Pipeline. The funnel depicts the order in which the entire proteome of Mtb is considered and analysed at different layers. 'A' refers to the systems level studies, which includes A1, for network analysis of the interactome; A2, for flux balance analyses of the reactome; and A3, for genome-scale essentiality data determined experimentally as reported by Sassetti et al [23]. Those proteins that passed these filters are indicated as 'A', and combined with the results of sequence analysis (B), to derive those that passed both filters (depicted as 'A&B'). These were then taken through Filter C, referring to the structural assessment filter, yielding the list of 622 proteins as the D-List (A&B&C). Further steps of filtering are indicated in the smaller funnel as E (expression under various conditions), F (non-similarity to anti-targets) and G (non-similarity to gut flora proteins). Those proteins that pass all the six levels of filtering (indicated as D&E&F&G) form the H-List comprising 451 targets. Additional filters I, J and K used for analysing the H-List are also indicated. Lists A', C' and E' refer to the set of proteins at A, C and E levels, respectively, that could not be analysed for lack of appropriate data. Lists A_X, B_X, C_X, E_X, F_X and G_Xrefer to sets of proteins that failed in that particular filter, but may have passed at other levels.