Analyzing Large Biological Datasets with an Improved Algorithm for MIC

2014/03/14 by Shuliang Wang, Wang, Shuliang, Yiping Zhao +1
Biochemistry, Genetics and Molecular Biology · Computer Science · #Computational Drug Discovery Methods #Machine Learning in Bioinformatics #Rough Sets and Fuzzy Logic #cs.CE #cs.DB

paper · pdf · doi:10.48550/arxiv.1403.3495

arxiv created 2014/03/14 · arxiv updated 2015/07/21

Abstract

A computational framework utilizes the traditional similarity measures for mining the significant relationships in biological annotations is recently proposed by Tatiana V. Karpinets et al. [2]. In this paper, an improved approximation algorithm for MIC (maximal information coefficient) named IAMIC is suggested to perfect this framework for discovering the hidden regularities between biological annotations. Further, IAMIC is the enhanced algorithm for approximating a novel similarity coefficient MIC with generality and equitability, which makes it more appropriate for data exploration. Here it is shown that IAMIC is also applicable for identify the associations between biological annotations.

Citations

• Equitability Analysis of the Maximal Information Coefficient, with Comparisons

Related

• Hierarchical Clustering Based on Mutual Information
• Information-theoretic Interestingness Measures for Cross-Ontology Data Mining
• The similarity metric
• Statistical Analysis based Hypothesis Testing Method in Biological Knowledge Discovery
• Effectively integrating information content and structural relationship to improve the GO-based similarity measure between proteins