Renormalised expression data from 1570 breast cancer related samples

An integration of several studies normalized to enable consistent cross-study comparisons and enrichment of the less frequent breast cancer types, enabling detection of new potential cancer driver genes. Raw Affymetrix CEL files were retrieved from the series GSE2034 GSE11121 GSE20194 GSE1456 GSE2603 GSE6532 GSE20437 GSE1561 GSE7390 GSE5847. Samples were then filtered with a three stage quality control resulting in the final dataset. These were subjected to RMA normalization and then placed on a log2 scale. Samples were stratified by levels of MYC expression with the top quartile defined as High and the bottom quartile defined as Low.