An Analysis of Breast Cancer Subtypes Using Homological Methods

In 2008, breast cancer caused 458,503 deaths worldwide in women according to The International Agency for Research on Cancer.  The four major subtypes of breast cancer are Normal-like, Luminal, Basal-like, and HER-2.  Each breast cancer subtype has a different prognosis and responds differently to chemotherapy.  We propose that we will be able to differentiate between different breast cancer subtypes by homological methods.   We used the ErbB signaling pathway network from the Kyoto Encyclopedia of Genes and Genomes along with actual cancer microarray data to determine if breast cancer subtypes differ topologically.  The networks that describe cellular pathways can be so large and complex that new mathematical methods must be developed to study them.  The approach that we used was to examine the ErbB signaling pathway network along with the cancer data to determine the topological properties of each cancer subtype.  Java programs that we have written, as well as previously existing software such as JPlex and Concom, were used to implement our methods. The preliminary results indicate that by our topological analysis we can distinguish between HER2 and Normal-like cancer subtypes. A computational method that can quickly determine different breast cancer subtypes based on microarray data will be valuable for both cancer treatment and cancer research.