Construct classification models for analysis of microarray dataset with gene ontology and OWL DL reasoner

This paper presents a novel method for microarray dataset analysis by incorporating the Gene Ontology to the construction of classification models to predict gene functions from its expression patterns in microarray dataset. Firstly from Gene Ontology build the hierarchical classification tree with an OWL DL reasoner and then figure out the subtree related to the microarray dataset. From the classification subtree, hierarchical classifiers can be constructed. The obtained hierarchical classification model represents the prediction ability of the dataset, which can direct the classification analysis objectively. A demonstration is given to show all the processes. Finally, we draw some conclusions and address our future works. The method can be generalized to similar scenarios to construct data analysis models aided by ontology.