Computational framework for streamlining the success of sequential antibiotic therapy

The authors present an open-source computational platform that leverages mathematical modelling of collateral sensitivity (i.e., when resistance to one antibiotic increases sensitivity to another) to design sequential antibiotic therapies. They formalise a framework that integrates experimental data (minimum inhibitory concentration changes, mutation profiles) into a network and dynamical model to predict which antibiotic sequences are likely to fail by driving multi-drug resistance. They validate the approach by demonstrating how certain sequential regimens (e.g., in chronic *Pseudomonas aeruginosa infections) would lead to treatment failure under the model. Overall, the tool offers a data-driven way to anticipate resistance evolution and streamline the selection of antibiotic sequences to prolong the efficacy of existing drugs.