Identifying High PPI Clusters by Integrating Data and Knowledge Sources

Developing computational methods to construct gene networks from time series profiles can help biologists generating and testing hypotheses to investigate the dynamics of complex regulatory systems. To tackle the problem of scalability, we develop a hybrid method by integrating data and knowledge so...

Full description

Saved in:
Bibliographic Details
Published inLecture notes in engineering and computer science Vol. 2235/2236; p. 1
Main Authors Lin, Chung-Hsun, Hsiao, Yu-Ting, Lee, Wei-Po
Format Journal Article
LanguageEnglish
Published Hong Kong International Association of Engineers 04.07.2018
Subjects
Clustering
Knowledge management
Similarity
Online AccessGet full text

Cover

More Information
Summary:Developing computational methods to construct gene networks from time series profiles can help biologists generating and testing hypotheses to investigate the dynamics of complex regulatory systems. To tackle the problem of scalability, we develop a hybrid method by integrating data and knowledge sources for network construction. Our approach includes a dimension reduction procedure to derive data features to calculate data similarity, and a knowledge mapping procedure to measure the semantic similarity between any two genes. A fuzzy gene clustering procedure is then performed. Experiments are conducted to evaluate the presented approach. The results show that our approach can produce meaningful clusters leading to better network construction performance.
ISSN:2078-0958
2078-0966