Intellectual engineering of alpha fetoprotein-arachidonic acid interactions to develop a novel and intelligent chemometric-amperometric biosensor based on exploiting first-order advantage: A new outcome of chemometry-electrochemistry engagement in bioanalysis

In this work, a novel amperometric biosensor was fabricated based on modification of a rotating glassy carbon electrode (GCE) with multiwalled carbon nanotubes-ionic liquid (MWCNTs-IL) and arachidonic acid (AA) (AA-MWCNTs-IL/GCE) for ultrasensitive and selective determination of alpha fetoprotein (AFP). The AFP was able to bind AA with high affinity which increased the steric hindrance at the biosensor surface and decreased reaching the probe molecules to the biosensor surface and caused the biosensor response to changed. Effects of experimental parameters on both structure and response of the biosensor were optimized by a central composite design (CCD). Under optimized conditions, first-order advantage was exploited from first-order amperometric data by modeling of them with PLS-1, rPLS, LS-SVM, PCR, CPR, RCR, BP-ANN, WT-ANN, PRM, DWT-ANN, RBF-ANN, and RBF-PLS to select the best algorithm to assist the biosensor for determination of AFP in blood samples having complex matrices. The results confirmed the RBF-ANN showed the best performance to assist the biosensor for determination of AFP in blood samples whose performance was comparable with liquid-phase binding assay using anion exchange chromatography as reference method. The biosensor developed in this work could be highly recommended as a reliable technique for bioanalyses towards determination of AFP.