Modeling and Optimization of Oligonucleotide-Based Nanobiosensor Using Artificial Neural Network and Genetic Algorithm Based Procedure

Developing a biosensor faces the different challenges for parameter optimization and calibration. In this study, a machine learning based approach is used to model and optimize the effective parameters of an electrochemical nanobiosensor based on thiolated probe-functionalized gold nanorods (GNRs) decorated on the graphene oxide (GO) sheet on the surface of a glassy carbon electrode (GCE). The response of the biosensor was considered as the output and eight effective factors including GO concentration, GNR concentration, probe concentration, probe time, MCH time, hybridization time, Oracet Blue (OB) concentration, and OB incubation time were used as inputs to train and model an artificial neural network. The experimental results demonstrate that the output of the developed model has an acceptable compatibility with the results obtained in the laboratory. The developed model is able to predict the output of the nanobiosensor with accuracy of 96.91% and the mean absolute percentage error (MAPE) value of 5.5090 %. Finally, genetic algorithm is used to find the optimum values of these parameters which yield the maximum value of the nanobiosensor output. The optimization results indicated that this method has better performance compared to the laboratory results and this method can be used for nanobiosensor design.