Computational Prediction of Genetic Drivers of Breast Cancer Metastases

点击翻译按钮获取中文摘要

Project Summary While metastasis is the main cause of cancer mortality, many of its molecular requirements remain unknown. Computational methods that trace the biological networks from upstream metastasis drivers to downstream effectors have the potential to identify new points of intervention for cancer therapies. Developing these types of methods requires individuals with expertise in statistical network models and deep engagement with cancer experimental systems. This fellowship will train the PI to predict the flow of both transcriptional and signaling information through biological networks, connecting these changes to phenotypic and behavioral consequences for cells and tissue derived from metastatic and non-metastatic breast cancer organoids. The Bader (computational) and Ewald (experimental) labs are jointly funded by the National Cancer Institute as a Cancer Target Discovery and Development (CTD2) Center focused on breast cancer metastasis. This center provides a uniquely powerful environment of mentorship, resources, and infrastructure that will enable the PI to use his formal training in statistical physics as the foundation for developing and applying new methods for computational oncology. Research will exploit three-dimensional organotypic cell culture and experimental methods motivated by population genetics to identify metastasis driver and effector genes in genetically engineered mouse models and in primary tumor specimens from an ongoing IRB-approved human subjects study. Genes identified by RNA- Seq will be analyzed with novel network perturbation theory to connect upstream drivers to downstream effectors. These inferred networks will in turn be connected to phenotypic and behavioral consequences for mammary organoids and tissues. Although outside the scope of this proposal, the JHU CTD2 Center has the mission and resources to validate findings with clinical potential for preventing or treating metastatic breast cancer, and potentially other invasive or metastatic cancers with similar molecular mechanisms. The PI’s background in biological and statistical physics, including computational methods, enables the mathematical and computational aspects of the proposed research. The fellowship will provide essential training that will permit the PI to lead independent research that combines physical sciences methods with experimental innovations and data-rich -omics measurements for cancer basic research and to aid therapeutic advances. The PI will have robust opportunities to collaborate with Hopkins and other institutions in future and will be an effective mentor for training computational oncology researchers in his own lab.