Abstraction of DNA Graph Structures for Efficient Enumeration and Simulation

ion of DNA Graph Structures for Efficient Enumeration and Simulation Ibuki Kawamata, Fumiaki Tanaka, Masami Hagiya Department of Computer Science, Graduate School of Information Science and Technology, University of Tokyo ibuki@is.s.u-tokyo.ac.jp, fumi95@is.s.u-tokyo.ac.jp, hagiya@is.s.u-tokyo.ac.jp Abstract— We propose a graph model of DNA molecules and an abstraction of that model for efficient simulation of molecular systems powered by DNA hybridization. In this paper, we first explain our DNA molecule model composed of graph data structures and highlight the problem of the large number of DNA structures that results. We then define an abstraction of the model, which focuses on local structures of DNA strands, and introduce reactions among the local structures. To verify the effectiveness of the abstraction, we develop simulators for the original and abstract models, and compare the number of structures generated by those simulators. Based on this research, computer-aided design of reaction systems that consist of biological molecules may become easier than conventional designs that rely on human trial and error. We propose a graph model of DNA molecules and an abstraction of that model for efficient simulation of molecular systems powered by DNA hybridization. In this paper, we first explain our DNA molecule model composed of graph data structures and highlight the problem of the large number of DNA structures that results. We then define an abstraction of the model, which focuses on local structures of DNA strands, and introduce reactions among the local structures. To verify the effectiveness of the abstraction, we develop simulators for the original and abstract models, and compare the number of structures generated by those simulators. Based on this research, computer-aided design of reaction systems that consist of biological molecules may become easier than conventional designs that rely on human trial and error.