Abstract—Fuzzy set theory has been widely used in the analysis of gene microarray data. However, due to noise and uncertainty inherent in microarray data, traditional fuzzy methods sometimes do not perform well. In this paper, we propose a type-2 fuzzy membership test (Type-2 FM test) for disease-associated gene identification on microarrays to improve traditional fuzzy methods. We apply this method on diabetes and lung cancer microarrays and make a comparison with traditional fuzzy methods. For diabetes data, we can identify 7 genes which have been confirmed to be related to diabetes treatment in the published literature and one more gene can be identified than original approaches. For lung cancer data, we can also identify 7 genes which have been confirmed to be associated with lung cancer treatment in published literature and the type-2 d-values are significantly different. The results show that our type-2 FM test performs better than traditional fuzzy methods when analyzing microarray data with similar expression values and noise.

Index Terms—type-2 fuzzy set; membership; microarray data; uncertainties.

Yan-Fei Wang is a PhD candidate in applied mathematics. He is now studying in Discipline of Mathematical Sciences, Queensland University of Technology, Brisbane, QLD, 4001 Australia (e-mail: yfmu@sina.com)
Zu-Guo Yu is a Professor in Mathematics at Xiangtan University of China and a Research Fellow in Mathematics at Queensland University of Technology, Australia. (corresponding to: 086-731-58298760; fax: 086-731-58293934; e-mail:yuzg@hotmail.com).

Cite:Yan-Fei Wang, Zu-Guo Yu, "A Type-2 Fuzzy Method for Identification of Disease-related Genes on Microarrays," International Journal of Bioscience, Biochemistry and Bioinformatics vol. 1, no. 1, pp. 73-78, 2011.