Emergence of drug resistance is a major problem in the treatment of many diseases including tuberculosis. To tackle the problem, it is essential to obtain a global perspective of the molecular mechanisms by which bacteria acquire drug resistance. Systems biology approaches therefore become necessary. This work aims to understand pathways to drug resistance and strategies for inhibition of the resistant strains by using a combination of experimental genomics and computational molecular systems approaches. Laboratory evolution of Mycobacterium smegmatis MC2 155 by treatment with isoniazid (INH), a front-line anti-tubercular drug, resulted in a drug-resistant strain (4XR), capable of growth even at about 10-times the minimum inhibitory concentration of the drug. Whole genome sequence of the 4XR was determined, which indicated only 31 variations in the whole genome, including 3 point mutations, 17 indels and 11 frame-shifts. Two mutations were in proteins required for the pharmacological action of the drug, albeit in regions distant from the drug binding site. The variations however were insufficient to explain the observed resistance to isoniazid. For a better understanding of the global changes associated with drug resistance, whole genome-wide gene expression data was obtained for the resistant strain and compared with that of the WT strain. 716 genes were found to be differentially regulated in 4XR, spanning different biochemical, signaling and regulatory pathways. From this, some explanations for the emergence of drug resistance were obtained, such as the up-regulation of the enzymes in the mycolic acid biosynthesis pathway and also of the drug efflux pumps. In addition, enrichment analysis indicated that up-regulated genes belong to functional categories of response to stress, carbohydrate metabolism, oxidation-reduction process, ion transport, signaling as well as lipid metabolism. The differential gene regulations seemed to be partially responsible for conferring the phenotype to the organism. Alterations in the metabolic pathways in 4XR were characterized using the phenotypic microarray technology, which experimentally scanned the respiratory ability of the resistant bacteria under 280 different nutrient conditions and 96 different inhibitors. Phenotypic gain, where the resistant strain grows significantly better than the wild type and phenotypic loss, where the growth of the resistant strain is compromised as compared to the sensitive strains were derived from the comparison of the phenotypic responses. Differences in survival ability and growth rates in different nutrient sources in the resistant phenotype as compared to the wild type were observed, suggesting rewiring in the metabolic network of the drug-resistant strain. In particular, the pathways of central carbon metabolism and amino acid biosynthesis exhibit significant differences. The strain-specific metabolic pathway differences may guide in devising strategies to tackle the drug-resistant strains selectively and in a rational manner. Scanning electron microscopy indicated the morphology of the drug-resistant strains to be significantly altered, as compared to the control drug-sensitive strain. It is well-known that isoniazid acts by inhibiting mycolic acid biosynthesis. The pathway turns out to be a target for many other anti-tubercular drugs also, since mycolic acids are major components of the cell wall. It is therefore important to understand what changes occur in the mycolic acid and the associated pathways in the drug-resistant variety so that strategies to tackle the latter can be chosen more judiciously. The lipidome of the cell wall was therefore quantitatively characterized by mass spectrometric analyses, which indeed confirmed that the 4XR strain has a significantly different composition profile. Among the six categories of lipids, the members of the glycerophospolipids category were abundant while the fatty acyls, polyketides and saccharolipids were lower in the 4XR strain as compared to the WT. The lipidomic data derived from the cell wall of INH-resistant strain shows that it results in the mycolic acid pathway function restoration, which would otherwise be lost upon drug exposure in the sensitive strain. Understanding the precise changes that occur in the lipidome in the drug-resistant strains is expected to be useful in developing new ways to tackle resistance. Next, to understand the implications of altered gene expression profiles, protein-protein interaction networks are constructed at a genome-scale that captures various structural and functional associations mediated by proteins in the mycobacterial cell. Using transcriptome data of 4XR, a response network is computed. Using an algorithm previously developed in the laboratory, the networks have been mined to identify highest differential activity paths and possible mechanisms that are deployed by the cells leading to drug resistance. Known resistance mechanisms such as efflux, cytochromes, SOS, are all seen to constitute the highest activities for achieving drug resistance in 4XR. Interestingly, such paths are seen to form a well-connected subnet, indicating such differential activities to be orchestrated. This clearly shows that multiple mechanisms are simultaneously active in the 4XR and may together generate drug resistance. Mechanisms of detoxification and antioxidant responses are seen to predominate in the 4XR subnet. Overall the analysis provides a shortlist of strategies for targeting the drug resistant strain. Next, the phenotypic microarray platform was used for screening for growth in Msm in the presence of various drugs. Data analysis and clustering resulted in identification of conditions that lead to phenotypic gain or loss in the 4XR as well as those that lead to differential susceptibility to various drugs. Drugs such as cephalosporins, tobramycin, aminotriazole, phenylarsine oxide, vancomycin and oxycarboxin were also found to inhibit growth in the resistant strain selectively. In other words, the 4XR is found to be collaterally sensitive to these drugs. The top-net formed by the highest differential activity paths, identified from the network described earlier has already indicated the involvement of proteins that generate antioxidant responses. Insights from the two methods, first from the targeted approach and second, from the phenotypic discovery approach were combined together to select only those compounds to which the 4XR strain was collaterally sensitive and targeted proteins responsible for antioxidant responses. These compounds are vancomycin, phenylarsine oxide, ebselen and clofazimine. These were further tested against the virulent M. tuberculosis H37Rv strain in a collaborator‘s laboratory. 3 of these compounds such as vancomycin, ebselen and phenylarsine oxide were found to be highly active in combinations with isoniazid against all tested Mtb strains, showed high levels of inhibition against H37Rv and 3 different single drug resistant, MDR and XDR strains. Moreover, they were observed to be highly potent when given in combinations. Clofazimine on the other hand, in combination with isoniazid showed activity but no significant synergy in the virulent drug-resistant strains of M. tuberculosis though synergistic to the sensitive strain. Thus, experiments with M. tuberculosis provide empirical proof that four different compounds, all capable of blocking antioxidant responses, are capable of inhibiting growth of single-, multiple- and extremely-drug-resistant clinical isolates of M. tuberculosis. Using transcriptome data from literature for M. tuberculosis exposed to six different drugs, similar drug specific response networks were constructed. These networks indicate differences in the cellular response to different drugs. Interestingly, the analysis suggests that different drug targets and hence different drugs could trigger drug resistance to various extents, leading to the possibility of prioritizing drug targets based on their resistance evolvability. An earlier study from the laboratory suggested the concept of target-co-target pairs, where-in the co-target could be a key protein in mediating drug resistance for that particular drug and hence for its target protein. Top ranked hubs in multiple drug specific networks such as PolA, FadD1, CydA, a monoxygenase and GltS, can possibly serve as co-targets. Simultaneous inhibition of the co-target along with the primary target could lower the chances of emergence of drug resistance. Such analyses of drug specific networks provide insights about possible routes of communication in the cell leading to drug resistance and strategies to inhibit such communication to retard emergence of drug resistance. Since mutations in the target proteins are known to form an important mechanism by which resistant strains emerge, an understanding of the nature of mutations in different drug targets and how they achieve resistance is crucial. Sequence as well as structural bases for the resistance from known drug-resistant mutants in different drug targets is deciphered and then positions amenable to such mutations are predicted in each target. Mutational indices of individual residues in each target structure are computed based on sequence conservation. Saturated mutagenesis is performed in silico and structural stability analysis of the target proteins has been carried out. Critical insights were obtained in terms of which amino acid positions are prone to acquiring mutations. This in turn suggests interactions that are not desirable, thus can be translated into guidelines for modifying the existing drugs as well as for designing new drugs. Finally, the work presented here describes application of the systems biology approaches to understand the underlying mechanisms of drug resistance, which has provided insights for drug discovery on multiple fronts though target identification, target prioritization and identification of co-targets. In particular, the work has led to a rational exploration of collateral drug sensitivity and cross-resistance of the drug-resistant strain to other compounds. Combinations of such compounds with isoniazid were first identified in the M. smegmatis model system and later tested to hold good for the virulent M. tuberculosis strain, in a collaborative study. The combinations were found to be active against three different clinical drug-resistant isolates of M. tuberculosis. Therefore, this study not only reveals the global view of resistance mechanisms but also identifies synergistic combinations of promising drug candidates based on the learnt mechanisms, demonstrating a possible route to exploring drug repurposing. The combinations are seen to work at a much reduced dosage as compared to the conventional tuberculosis drug regimens, indicating that the toxicity and any associated adverse effects may be greatly reduced, suggesting that the combinations may have a high chance to succeed in the next steps of the drug discovery pipeline.