We selected 407 patients with both tumor samples and matched normal blood samples from the TCGA breast cancer dataset. For each of these samples, we obtained the germline genotypes from the normal blood sample and somatic copy number, methylation, mRNA gene expression measures from the matched tumor sample (Table S1, related to Figure 1). Ancestry was inferred using genotype data resulting in 273 cases of European ancestry. In the analyses below, 171 ER-positive cases and 48 ER-negative cases (as determined by ESR1 expression levels) were used (Methods).
