Parameter estimation of probabilistic Boolean control networks: An optimization-based approach

Abstract

This paper addresses the parameter estimation of probabilistic Boolean control networks (PBCNs) that exhibit unknown and time-varying switching probability distributions (SPDs). We begin by transforming PBCNs with time-varying SPDs into an algebraic form within the framework of algebraic state-space representation (ASSR). In contrast to traditional PBCNs, our investigation incorporates uncertainty through variable switching probabilities. We propose an optimization-based approach for estimating the unknown and time-varying SPD of PBCNs. While existing methods typically focus on estimating Boolean networks (BNs) states, our approach targets the more challenging task of estimating time-varying SPDs. We validate the effectiveness of the proposed method using a simplified apoptosis network model.