Bioinformatics study of the molecular features of breast cancers

A major application of current functional genomics is to study the underlying molecular properties of diseases like cancer by profiling the expression of thousands of genes. Pattern recognition techniques have been widely used for such applications. Based on microarray gene expression data of breast cancer samples, we conducted a series of analysis on various features of breast cancer patients. The studied features include ER, Her2, grade and lymph node status. The observations from the data have provided new understanding to the biology behind discimination between various breast cancer features. Especially, a putative pathway of microRNA regulation on the differentiation of breast tumors has been proposed based on the intergrated investigation on gene expression, microRNA targeting and gene set analysis.