As Mycobacterium tuberculosis has become drug resistant, we need to
design new anti-tuberculosis lead compounds. This review has focused on the
application of various chemoinformatics methods that can be used in an attempt to
search for potent and selective inhibitors against M. tuberculosis. One of the
“rational” approaches towards designing such novel anti-bacterials is to develop and
deploy QSAR of similar drug-like compounds, that helps in implementing and
improvising the techniques and shares some newly identified potential anti-TB drug
candidates. Here, we have also mentioned about some of the QSAR models
developed and validated by various groups as well as our team for several
derivatives showing anti-tuberculosis activity viz. fluoroquinolones, quinoxaline and
nitrofuranyl amide derivatives etc. As the calculation of diverse physicochemical
properties for such huge number of compounds is time consuming and also not costeffective,
we have utilized molecular descriptors for regression modeling. Among
different types of descriptors, the study has also been extended to understand the
influence of each class of molecular descriptor for predicting structure-activity
relationships, and the results indicate the preeminence of topological descriptors
over other descriptor lessons. The methodologies described in this review are non
specific and applicable to other syndromes also.
Keywords: Fluoroquinolone, molecular descriptors, molecular docking,
molecular similarity, nitrofuranyl amide, QSAR, quinoxaline, statistical
regression models, tuberculosis, virtual screening.
