A Hybrid Analytical and Machine Learning Model of BioFETs for the Detection of Peptides

BioFETs have been employed to detect a wide range of biological analytes by sensing changes in surface potential, translating them into electrical response. While analytical models such as the Gouy–Chapman–Stern and site-binding models successfully describe surface charge and protonation equilibria of these analytes, they do not capture the sequence-dependent chemistry, steric hindrance, and conformational variability intrinsic to peptides and other biomolecules that define the sensor response. We propose a hybrid molecular-analytical model to address these limitations, augmenting previous BioFET simulation frameworks with modelled molecular parameters including dissociation constants and topological surface areas. While this work focuses on peptides, including post-translational and synthetic modifications, the framework extends to diverse analytes and immobilization chemistries relevant to biosensing. We validate the model against experimental values and molecular modelling studies, confirming the physical fidelity of our hybrid model. Simulations with three representative peptides demonstrate the model's capacity to resolve fine chemical variations, providing a practical tool to guide experimental design and contribute to the development of BioFETs.