Mining fuzzy association patterns in gene expression data for gene function prediction

The development in DNA microarray technologies has made the simultaneous monitoring of the expression levels of thousands of genes under different experimental conditions possible. Due to the complexity of the underlying biological processes and also the expression data generated by DNA microarrays are typically noisy and have very high dimensionality, accurate functional prediction of genes using such data is still a very difficult task. In this paper, we propose a fuzzy data mining technique, which is based on a fuzzy logic approach, for gene function prediction. For performance evaluation, the proposed technique has been tested with a genome-wide expression data. Experimental results show that it can be effective and outperforms other existing classification algorithms. In the separated experiments, we also show that the proposed technique can be used with other existing clustering algorithms commonly used for gene function prediction and can improve their performances as well.