NSF funded project. Grant number 1217324

Breast cancer is a very heterogeneous disease that in the US affects 200,000 people per year. It is clinically classified according to certain parameters such as Stage and grade. In addition, molecular phenotyping has allowed the categorization of tumors using expression of surrogate markers of Estrogen and Progesterone Receptors (ER) and (PR) and of the gene ERBB2 (also known as HER2). Most microarrays studies focus on finding specific genes that are either amplified or deleted and whose expression is correspondinly changed. We use computational homology to test for differences in the copy number (and gene expression) topography of each chromosomes. Therefore we interrogate not to single probes/genes but to combinations of nearby probes using techniques from algebraic homology.