Chemogenomic model identifies synergistic drug combinations robust to the pathogen microenvironment
Fig 2
Predicting multi-drug combinations using MAGENTA.
(a) Schematic workflow of MAGENTA approach. MAGENTA takes drug chemogenomic profiles and interactions among drugs as input. This is used to train a Random Forest Model which can predict synergy and antagonism among pair-wise or higher-order combinations of new drugs given their chemogenomic profiles. (b) All three-way interactions among 8 antibiotics are represented as 3D heat-map. Blue, white or red boxes correspond to synergistic, additive or antagonistic three-way combinations. Among 56 combinations, only Azi+Min+Rif and Min+Cip+Rif exhibit strong synergy. The three-way interactions among 8 antibiotics are also represented as layers for maximum visibility. The dotted lines depict the interaction between nitrofurantoin, minocycline and chloramphenicol. (c) Scatter plot comparison of MAGENTA 3-way interaction predictions and experimental measurements demonstrate that MAGENTA can robustly identify 3-way antibiotic synergy and antagonism (rank correlation R = 0.57, p = 5 x10-6).