Using k-shortest paths to predict unknown metabolic pathways

• Main Authors: Yu-Fu Lin, 林育甫
• Other Authors: Hei-Chia Wang
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/62900728120730850879

碩士 === 國立成功大學 === 資訊管理研究所 === 94 === 　　In the post-genomic era, we not only hope to understand the information of biological sequence, and further, a greater challenge is to explore the metabolic pathway formed by interactive reactions between genes. 　　Since the computer technique improves rapidly, the computer generated biological data and experimental results increased exponentially. There are more and more public databases provided free and huge amount of biological data for biological researchers. In recent years, many researchers utilize the data source provided by there public databases, to construct metabolic pathway, and to infer the metabolic pathway by graph theory. But these researches usually ignored the choice of start and end nodes. The biological researchers must input these nodes manually, but it may cause the fault of inferring results. Hence, the method we propose in this research is to retrieve the nodes of reference pathway of KEGG pathway database, to ensure the nodes we retrieved are all in the similar function. We utilize k-shortest paths to infer the metabolic pathway in the metabolic pathway network which constructed by compounds and reactions. Finally, the candidated metabolic pathways we inferred are provided for the biological researcher to further analyze.