Human cancers are complex diseases that present vast heterogeneity, leading to widely different survival outcomes. Thus actionable and robust predictive models for cancer prognosis are much needed for more personalized treatment and disease management. However, There has been severely lacking of therapeutically actionable computational methods, which explicitly model patient survival difference as the objective while integrating multi-dimension high-throughput genomics data. To solve this issue, we propose a novel actionable framework for caner drug repurposing, called DR. EPS (Drug Repurposing for Extended Patient Survival). Towards this goal, we will develop the following strategies: (1) Constructing and validating a new class of hybrid-learning based, multi-omics prognosis modeling approach, using seven diverse liver cancer population cohorts. (2) Developing and experimentally validating an actionable computational drug- repurposing framework to improve the survival of high-risk liver cancer patients, using big sets of pharmacogenomics and pharmacogenetics data. (3) Building a user-friendly webtool that accelerates such bench-to-bed transition for liver cancer treatment. We expect that this project will be groundbreaking in many aspects, including building new prognostic models on multi-omics data sets, identifying new repurposed drugs to extend high-risk liver cancer patient life-spans, and providing a first-hand drug reposition resource for liver cancer therapeutics research community.
The goal of this proposal is to develop an actionable drug-repurposing framework, in order to extend the survival time of high-risk liver cancer patients. We will first build and validate novel and robust prognostic computational models that integrate multi-omics cancer data, then use signatures obtained from the high-risk survival group are to identify repurposed drug candidates. This study is expected to enhance more personalized management and accelerate research in drug repurposing and reposition in liver cancers.