Although feature selection has proven effective in sample class prediction, it is not adequate for identifying leads for potentially useful biomarkers by high-throughput biological data analysis. The large number of equally good predictive sets and the disparity among them reveals the gap between feature selection and biomarker identification. We propose to bridge this gap by a new data mining task, feature clus- ter selection, which aims to select and group all relevant features in a data set into a small number of coherent clus- ters. We provide both theoretical framework and empirical formulation for the new problem, and propose the 3M algo- rithm. Experiments on microarray data show that the algo- rithm can select highly predictive representative gene sets and discover gene clusters with statistical significance.