Authors : Maria Joseph and Sreeja Ashok
Abstract: Bioinformatics is an important area in which computing techniques can be applied for efficient data analysis and for mining meaningful patterns. The organization, analysis and interpretation of data are the major challenges faced by biologists when dealing with large amount of heterogeneous and complex data. Unsupervised learning techniques are widely used for data reduction and pattern extraction for in-depth analysis and knowledge discovery. Graph clustering is a more suitable approach, since, the interactions of the biological components can be effectively demonstrated by networks. The complexity of the graphs can be reduced by extracting highly significant edges instead of focusing on all edges that represents the association between data objects. This study compares and evaluates the significance of different Minimum Spanning Tree algorithms (MST) as a preprocessing step for community detections in biological data. Multiple algorithms were reviewed and compared and the process performance is compared with benchmark community detection algorithms.
Maria Joseph and Sreeja Ashok, 2017. Minimum Spanning Tree Based Community Detection for Biological Data Analysis. Journal of Engineering and Applied Sciences, 12: 5452-5456.