State-space modeling for dynamic response of graphene FET biosensors

Abstract Graphene field effect transistor (G-FET) biosensors exhibit high sensitivity owing to their high electron/hole mobilities and unique 2D nature. However, a baseline drift is observed in their response in aqueous environment, making it difficult to analyze their response against target molecules. Here, we present a computational approach to build state-space models (SSMs) for the time-series data of a G-FET biosensor; the approach helps separate the response against target molecules from the baseline drift. The charge neutral point of the G-FET sensor was continuously measured while sensing target molecules. The obtained time-series data were modeled using the proposed SSMs. The model parameters were estimated through Markov chain Monte Carlo methods. The SSMs were evaluated using the widely-applicable Bayesian information criterion. The SSMs well fitted the time-series data of the G-FET biosensor, and the sensor response to target molecules was extracted from the baseline-drift data.