Abstraction of graph-based models of bio-molecular reaction systems for efficient simulation

Ibuki Kawamata, Nathanael Aubert, Masahiro Hamano, Masami Hagiya

研究成果: 書籍の章/レポート/Proceedings会議への寄与査読

5 被引用数 (Scopus)

抄録

We propose a technique to simulate molecular reaction systems efficiently by abstracting graph models. Graphs (or networks) and their transitions give rise to simple but powerful models for molecules and their chemical reactions. Depending on the purpose of a graph-based model, nodes and edges of a graph may correspond to molecular units and chemical bonds, respectively. This kind of model provides naive simulations of molecular reaction systems by applying chemical kinetics to graph transition. Such naive models, however, can immediately cause a combinatorial explosion of the number of molecular species because combination of chemical bonds is usually unbounded, which makes simulation intractable. To overcome this problem, we introduce an abstraction technique to divide a graph into local structures. New abstracted models for simulating DNA hybridization systems and RNA interference are explained as case studies to show the effectiveness of our abstraction technique. We then discuss the trade-off between the efficiency and exactness of our abstracted models from the aspect of the number of structures and simulation error. We classify molecular reaction systems into three groups according to the assumptions on reactions. The first one allows efficient and exact abstraction, the second one allows efficient but approximate abstraction, and the third one does not reduce the number of structures by abstraction. We conclude that abstraction is a useful tool to analyze complex molecular reaction systems and measure their complexity.

本文言語英語
ホスト出版物のタイトルComputational Methods in Systems Biology - 10th International Conference, CMSB 2012, Proceedings
ページ187-206
ページ数20
DOI
出版ステータス出版済み - 2012
イベント10th International Conference on Computational Methods in Systems Biology, CMSB 2012 - London, 英国
継続期間: 2012 10月 32012 10月 5

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
7605 LNBI
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

会議

会議10th International Conference on Computational Methods in Systems Biology, CMSB 2012
国/地域英国
CityLondon
Period12/10/312/10/5

フィンガープリント

「Abstraction of graph-based models of bio-molecular reaction systems for efficient simulation」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル