Offline Fault Detection in Gene Regulatory Networks using Next-Generation Sequencing Data

In a previous contribution, a method was proposed for on-line fault detection in Boolean gene regulatory networks based on the Boolean Kalman Filter and hypothesis testing on next-generation sequencing (NGS) data. In this paper, we present a related methodology, which differs in two main points: (1) it uses the Boolean Kalman Smoother on the entire data set stored off-line; (2) it employs a maximum-likelihood approach for simultaneous change detection and categorization. The efficacy of the proposed methodology is assessed using a p53-MDM2 negative feedback loop Boolean network with stuck-at faults that model molecular events commonly found in cancer.