Breast cancer patients with deficiencies in the homologous recombination (HR) pathway are sensitive to poly (ADP-ribose) polymerase (PARP) inhibitors and platinum-based chemotherapy. Various methods have been developed to detect HR deficiency, many of which rely on genomic features indicative of impaired HR machinery. However, genomic-based approaches cannot distinguish between current and past HR deficiency. In contrast, the transcriptome can capture the current state of homologous recombination and may be more effective than genome-based methods in reflecting the dynamic nature of the HR pathway in cancer. Recently, the clinical utility of transcriptional classifiers has been demonstrated for identifying HR-deficient (HRD) prostate cancers, but there remains a need for robust and widely-applicable transcriptome methods for detecting HR deficiency in breast cancer. In this thesis, I developed a 153-gene transcriptional signature for detecting HR deficiency in breast cancer patients, allowing identification of individuals whose cancers may be sensitive to PARP inhibitors. This signature offers an advantage over existing models due to its reduced feature set and ability to generalize across different subtypes of HRD breast cancer.