In this paper, we propose two fuzzy data mining approaches for microarray analysis, namely Fuzzy Associative Gene Expression (FAGE) and Ripple Effective Gene Expression Rule (REGER) algorithms. Both of them first transform microarray data into fuzzy items, and then use fuzzy operators and specially-designed data structures to discover the relationships among genes. Through the proposed algorithms, a novel pattern named Ripple Pattern is discovered that indicates the genes active at the same time with their linguistic terms being monotone increasing or decreasing. The experimental results show that the proposed algorithms are effective in discovering novel and useful rules from microarray data.