The individual response to therapeutic treatments is likely to be a complex trait that is controlled by various genetic and non-genetic factors. In an ideal world of personalized medicine, the information about a patient's genetic make-up or gene expression profile would be considered by physicians together with other clinical information (e.g., age, gender) to tailor medical care for both maximizing effective therapy and avoiding adverse effects. The challenge for personalized medicine is particularly urgent for cancer chemotherapy. Clinically, anticancer drugs often present a narrow therapeutic index, indicating that small changes in dosage could cause severe toxic response (e.g., neurotoxicity and nephrotoxicity) [1,2] with the extreme end of complication resulting in fatality. Therefore, understanding the comprehensive relationships between genetic/non-genetic factors and drug response is a critical step toward the realization of personalized medical care in clinical oncology.
During the past decade, high-throughput technological platforms have become available for comprehensively and quantitatively profiling genetic variation and molecular targets in cells. For example, the advances in microarray (e.g., the Affymetrix GeneChip and the Illumina BeadChip platforms) and RNA-sequencing (e.g., the Illumina Genome Analyzer) technologies have allowed the profiling of the entire transcriptome (i.e., mRNA-level expression) of cells including both common and rare transcripts, as well as transcript variants (i.e., transcript isoforms). For genetic variation, large research efforts such as the International HapMap Project [3,4] and the 1000 Genomes Project [5] have made publicly available a detailed map of human genetic variation across major populations including Asians, Europeans and Africans. Taking advantage of these technological and research advances, cell-based pharmacogenomic studies have begun to identify genetic determinants that are responsible for drug response [6,7].