Representing Genetic Networks as Labeled Hybrid Petri Nets for State Space Exploration and Markov Chain Analysis


This paper presents the bachelor’s thesis of Curtis Kendall Madsen which can be broken down into the following three goals. The first goal of this project is to develop a way to convert genetic networks into logical models. Once this is done, finding the state graph of these nets and performing Markov chain analysis on them can provide researchers with insight into the reachability of the states in the original network. Therefore, the second goal of this project is to develop an automated tool that can perform state space exploration of a logical model, and the third goal is to implement a Markov chain analyzer for the stage graph. For the logical representation of genetic networks, the conversion method uses labeled hybrid Petri nets (LHPNs) because they are designed for modeling logic while still allowing for important information required by Markov chain analysis such as transition rates to be included in the model. This conversion method is automated and is integrated into the iBioSim program allowing users to transform a Genetic Circuit Model (GCM) into an LHPN with just a click of a button. Also, iBioSim now includes the LHPN file type in its file tree so users can view and edit LHPNs once conversion is complete. In addition, a method for performing state space exploration on an LHPN allows the user to view the state graph using Graphviz’s Dotty tool.

Curtis Madsen
Curtis Madsen
Sandia National Laboratories, R&D S&E, Computer Science