Predicting the mechanism of action of antituberculosis agents using chemical global positioning system – natural product
Muaaz Mutaz Alajlani, Anders Backlund · Planta Medica · 2016
Tuberculosis is a highly infectious disease and one of the world's most serious public health threat. To achieve global control of this urgency, there is a pressing need for new TB drugs with more favourable characters in terms of pharmacokinetic properties, shorter treatment period, toxicity reduction, targeting MDR and XDR, and possibility of co-administration with HIV medications. Mechanism of action has never been used as a screening criterium due to complex, difficult and cost ineffective screening methods. In this presentation a novel perspective using chemical property space is proposed in predicting mechanism of action for compounds libraries at earlier stages. Different established mechanisms for 42 compounds antituberculosis were interpreted to three dimensional chemical property space by calculating principle component values of 35 physicochemical properties provided by chemical position system- natural product (ChemGPS-NP) [1]. These projections have formed a specific chemical property space for each mechanism (figure below). A total of 7435 compounds (GSK set 776, NIAID Set 3779 and GVK Bio Set 2880) with antituberculosis activity were assigned to different mechanisms by measuring the Euclidian distances making it possible to predict mechanism of action. The results were made possible to position large data sets of antituberculosis for upscale prediction and even in many case provided an insights on potential unique positions or mechanisms that did not fit the usual mechanisms in the chemical property space of antituberculosis compounds. Chemical property space of antituberculosis compounds with different mechanisms of actions shown as with colored clusters mechanism of actions (Yellow spheres as Nucleic acids synthesis inhibition, Green spheres as protein synthesis inhibition, pink spheres as cell wall synthesis inhibition and Red sphere as unique anti-ATP) projected by ChemGPS-NP where PC1 (x = red), PC2 (y = yellow) and PC3 (green).