Executive Summary : | Inter- and intra-tumoral heterogeneity in cancer treatment presents significant challenges, leading to differential drug responses among patients. The effectiveness of many drugs is limited to a subset of patients, causing potential consequences like treatment failure and adverse drug events. To achieve personalized cancer treatment, researchers need to understand how cancer patients respond to different anticancer drugs based on their genetic and clinical characteristics. To do this, researchers need to uncover precise patterns of drug responses through comprehensive patient drug screening. However, the feasibility of treating a vast population of cancer patients poses significant limitations. This project aims to provide an integrated computational tool to explore the relationships between drug response phenotypes and related molecular information and survival analyses based on patients' clinical outcomes. It incorporates biological knowledge, such as pathway and interactome data, during the processing of genomic features. The project proposes a biclustering algorithm incorporating biological knowledge to select high-weighted features from each of the omics data to get a single multi-omics representation. This approach aids in deciphering the complexity of cellular systems and provides valuable insights for drug response prediction, biomarker identification, and the development of targeted therapies. In conclusion, the development of this integrated computational tool represents a significant step towards achieving better cancer treatment and increasing survival rates for cancer patients. |