A novel amperometric biosensor for multi-enzymatic biosensing of triglycerides

This work has been focused on developing a novel biosensor assisted by multivariate calibration methods to determine triglycerides (TGs, triacetin, tributyrin, tricaproin, tricaprylin, and tricaprin) in lyophilized serum samples. To achieve this goal, a bare glassy carbon electrode (GCE) was modified and used as the biosensing platform. To increase the sensitivity of the developed method, hydrodynamic methods were used to calibrate the biosensor response. To increase the selectivity of the developed biosensor, it was assisted by partial least squares-1 (PLS-1), radial basis function-PLS (RBF-PLS), and RBF-artificial neural network (RBF-ANN) for exploiting first-order advantage. After characterization of the modifications, the first-order advantage was exploited to increase the selectivity of the method by building a multivariate calibration set in a pre-analyzed lyophilized serum with different TGs concentrations, which were chosen according to the individual calibration curve. Calibration models were then built in the same pre-analyzed lyophilized serum and analyzed by PLS-1, RBF-PLS, and RBF-ANN. Therefore, their performances were examined to predict concentrations of a validation set. The results confirmed the successfulness of the calibration model developed by RBF-ANN. Finally, it was used to analyze two serum samples, and the results demonstrated that the method was successful because its results were compared with a reference method.