Fractional-Order Modeling and Neural Network Simulation of NK Cell-Oncolytic Virus Interactions

Oncolytic virotherapy (OV) uses viruses to selectively destroy cancer cells, but the role of natural killer (NK) cells ‎in this process remains unclear. This study develops a discrete fractional-order model to examine NK cell activity ‎and identify activation conditions. Existence of the model solution is established via fixed-point theorems, and ‎the basic reproduction number (R_0) is derived using the next-generation matrix method with sensitivity analysis ‎to determine key parameters. Stability of equilibria is analyzed, and numerical simulations reveal how NK cell ‎activation affects interactions among cancer cells, infected cells, and viruses. Additionally, an artificial neural ‎network (ANN) trained with the Levenberg-Marquardt algorithm efficiently approximates the model's dynamics, ‎showing high accuracy (low MSE) and providing a rapid, reliable tool for predicting complex biological behaviors.