The data presented here come from a study of gene expression changes in ovarian cancer. The current standard treatment for ovarian cancer consists of surgery, followed by either carboplatin and paclitaxel or carboplatin alone. This approach, however, is not effective for all patients. The goal of this study was to identify genes and pathways associated with drug response. To identify such genes, the investigators implanted ovarian cell lines into adult mice and allowed the tumors to grow for 2 months, at which point one of three treatments (carboplatin, carboplatin + paclitaxel, or control) was administered to each mouse. At various time points ranging from 0 to 14 days following the initiation of treatment, the mice were sacrificed, at which point the investigators measured the size of the tumor as well as gene expression in the cancerous tissue.
Our analysis here concentrates on relative tumor volume (RTV) as the outcome variable.
For this study, there were 34,694 features with expression data and a sample size of 101 mice.
y: Relative tumor volume. Measurements on on the log2 scale so that y=1 means that the tumor has doubled in size since baseline. By definition, y=0 for all samples taken at day 0.X: Gene expression measurementssData: Data frame containing additional information about the experimental conditions for each sample.
Treatment: Treatment received (Carbo, CarboTax, Control)Day: Days since beginning of treatmentfData is provided that contains additional annotation information for each feature.Koussounadis A, Langdon SP, Harrison DJ and Smith VA (2014). Chemotherapy-induced dynamic gene expression changes in vivo are prognostic in ovarian cancer. British Journal of Cancer, 110: 2975-2984.