DescriptionSuccessful cancer treatment is based on our understanding of a number of biological considerations such as its mechanisms for survival, evasion of tumor suppressor programs, and proliferation. Unfortunately, cancer evolution is often chaotic and a single tumor may exhibit many different methods for achieving its goals, such as direct mutation of tumor suppressor genes, over-expression of genes which target tumor suppressors, or both. With that in mind, it is crucial for clinicians and researchers to be able to distinguish the properties of each tumor and identify similarities between them, so that broad-impact treatments can be devised. Recently, a number of advances have been made which allow researchers to gather more detailed information in a high-throughput manner on the behavior of individual tumors. Where once only gross gene expression information could be gleaned using a microarray chip, now sequencing technology enables us to understand what individual isoforms of genes are being expressed, and in what abundance. Sequencing technology advances have also enabled us to find novel sites of expression on the genome which do not correspond to known proteins, and in fact provide evidence of a new class of large non-coding RNA molecules with functional consequences for cancer tumors. In this thesis, we present novel methodologies for the identification of alternative transcript as well as non-coding RNA usage in subgroups of breast cancer tumors using data from next generation transcriptome sequencing. Using these methods, we have identified genes which are differentially spliced between breast cancer tumors belonging to estrogen positive (ER+) and negative (ER-) sets, as well as in novel subgroups, and validated the existence of these transcripts in tumor tissue RNA using RT-PCR. Additionally, we present evidence of non-coding RNA transcripts which are aberrantly expressed based on estrogen status, and validate these in a similar way. These discoveries and new methodologies will help elucidate the biological differences between these subgroups of breast cancer, and will assist ongoing research into transcriptome abnormalities in other cancers as sequencing data become available.