Automated Discovery of Long Intergenic RNAs Associated with Breast Cancer Progression.

摘要

Of the 40,000 women in the United States that die from breast cancer every year, almost half of them might have once thought they were cured. While treatment of early stage breast cancer can be curative, up to 30% of node- negative and 70% of nodepositive breast cancers will relapse. Therefore, risk stratification and surveillance of breast cancer is of paramount importance. Prognostic markers such as estrogen and progesterone receptor, ERBB2, and comedo necrosis currently guide clinical decisions, but these markers often fail to estimate the true risk of relapse for many patients. This results in frequent overtreatment of indolent cancer and undertreatment of high-risk disease. Over the past decade gene expression microarrays have facilitated the development of prognostic tests, but Next Generation Sequencing of cancer transcriptomes (RNA-seq) technologies can provide more information at higher accuracy. However, current RNA-seq analysis tools cannot resolve a significant portion of the data emanating from cancer transcriptomes, and it is precisely this data that could lead to the discovery of novel genetic aberrations and/or therapeutic targets. Therefore, we hypothesize that transcriptome sequencing can elucidate novel transcriptional aberrancies in breast cancer that may affect or predict disease prognosis. We will test this hypothesis with the following specific aims: (1) we will employ a combined alignment and assembly approach to increase the sensitivity of current analysis methods, (2) use this approach to detect novel transcriptional aberrancies in breast cancer, and (3) predict functional relationships between novel transcripts and known prognostic markers. Methods to detect and characterize these transcripts could lead to discovery of oncogenic mutations, delineate new molecular subtypes of breast cancer, and/or identify novel therapeutic targets for breast cancer treatment.