Computer-Aided Design of Eco-Friendly Imprinted Polymer Decorated Sensors Augmented by Self-Validated Ensemble Modeling Designs for the Quantitation of Drotaverine Hydrochloride in Dosage Form and Human Plasma.

Abstract

Computationally designed molecular imprinted polymer (MIP) incorporation into electrochemical sensors has many advantages to the performance of the designed sensors. The innovative self-validated ensemble modeling (SVEM) approach is a smart machine learning-based (ML) technique that enables the design of more accurate predictive models using smaller data sets. The novel SVEM experimental design methodology is exploited here exclusively to optimize the composition of four eco-friendly PVC membranes augmented by a computationally designed magnetic molecularly imprinted polymer to quantitatively determine drotaverine hydrochloride (DVN) in its combined dosage form and human plasma. Furthermore, the application of hybrid computational simulations such as molecular dynamics and quantum mechanical calculations (MD/QM) is a time-saving and eco-friendly provider for the tailored design of the MIP particles. Here, for the first time, the predictive power of ML is assembled with computational simulations to develop four PVC-based sensors decorated by computationally designed MIP particles using four different experimental designs known as central composite, SVEM-LASSO, SVEM-FWD, and SVEM-PFWD. The pioneering AGREE approach further assessed the greenness of the analytical methods, proving their eco-friendliness. The proposed sensors showed decent Nernstian responses toward DVN in the range of 58.60-59.09 mV/decade with a linear quantitative range of 1 × 10-7 - 1 × 10-2 M and limits of detection in the range of 9.55 × 10-8 to 7.08 × 10-8 M. Moreover, the proposed sensors showed ultimate eco-friendliness and selectivity for their target in its combined dosage form and spiked human plasma. The proposed sensors were validated in accordance with International Union of Pure and Applied Chemistry (IUPAC) recommendations, proving their sensitivity and selectivity for drotaverine determination in dosage form and human plasma. This work presents the first ever application of both the innovative SVEM designs and MD/QM simulations in the optimization and fabrication of drotaverine-sensitive and selective MIP-decorated PVC sensors.