Classifying subtypes of acute lymphoblastic leukemia using silhouette statistics and genetic algorithms

► Classifying heterogeneous cancer types has become a challenging topic in bioinformatics. ► We classify six subtypes of acute lymphoblastic leukemia with gene expression profiles. ► Wrapper-based gene selection methods are useful for pattern recognition in classifying microarray data. ► We introduce the new application of silhouette statistics for classification quality analysis.