Accelerating the design of synthetic biological circuits requires expanding the currently available genetic toolkit. Although whole-cell biosensors have been successfully engineered and deployed, particularly in applications such as environmental and medical diagnostics, novel sensing applications necessitate the discovery and optimization of novel biosensors. Here, we address this issue of the limited repertoire of biosensors by developing a data-driven, transcriptome-wide approach to discover perturbation-inducible genes from time-series RNA sequencing data, guiding the design of synthetic transcriptional reporters. By combining techniques from dynamical systems and control theory, we show that high-dimensional transcriptome dynamics can be efficiently represented and used to rank genes based on their ability to report the perturbation-specific cell state. We extract, construct, and validate 15 functional biosensors for the organophosphate malathion in the underutilized host organism Pseudomonas fluorescens SBW25, provide a computational approach to aggregate individual biosensor responses to facilitate enhanced reporting, and exemplify their ability to be useful outside the lab by detecting malathion in the environment. The library of living malathion sensors can be optimized for use in environmental diagnostics while the developed machine learning tool can be applied to discover perturbation-inducible gene expression systems in the compendium of host organisms.