AGI has used gene expression signatures to target the creation of antibody tools to improve classification and treatment of cancer.

We began by targeting antibody production to genes identified by gene expression patterns that distinguish novel subtypes of cancer. We generated and screened hundreds of new antibodies to identify sets of quality immunohistochemistry reagents specific to carcinomas arising from different tissue origins. We believe that these antibody sets distinguish a previously unappreciated spectrum of biologic diversity within each cancer type which likely could improve clinical development and targeting of chemotherapeutics. We call these antibody sets "Panels of Diversity." These panel are our starting tools to discover biomarkers for prognosis and to predict response to chemotherapeutics.

	Figure. This figure depicts the gene expression data [Sorlie et al. PNAS 14:8418(2003)] from a group of breast cancer patients. The data has been organized so that genes with similar expression pattern across the patients are grouped together. Adjacent to the expression data is a bar graph depicting the number of antibodies targeted to genes within each gene expression delineated subgroup. Many antibodies were targeted to each subgroup to identify the most robust immunohistochemistry reagents for inclusion in AGI's Panel of Diversity.