Interrogating Metastatic Prostate Cancer Biological and Clinical Trajectories Through Integrative Data Science
Abstract
Multiple research efforts have defined the molecular landscape of metastatic prostate cancer (MPC) and led to immediate clinical impact, including germline genetic testing and clinical studies of several treatments such as PARP inhibitors, immunotherapies, and PI3-kinase inhibitors. However, each of these studies remain limited in their ability to integrate combinatorial molecular findings with detailed clinical data for the translational MPC discovery. Harmonization across various studies remains computationally challenging due to multi-modal data types and insufficient compute power. In addition, these studies have not leveraged new profiling technologies and computational advances in artificial intelligence to learn meaningful representations of complex molecular and phenotypic features. To overcome these critical challenges, we have created a multidisciplinary and collaborative data science program to integrate molecular properties, transcriptional programs, and phenotypic features to determine clinical trajectories in MPC. We will harmonize clinical cohorts representative of MPC patient diversity and apply biologically guided interpretable deep learning methodology to understand how genomic features interact in promoting MPC biology to reduce disparities in prostate cancer. We will also reveal interacting tumor and immune transcriptional programs in MPC using a harmonized interactive cloud-based framework to provide pathways toward novel therapeutic development in MPC.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 28, 2022
- Source ID
- W81XWH2210085
Entities
People
- Eliezer Van Allen
Organizations
- Dana–Farber Cancer Institute
- United States Army