Resolving phenotypic overlap in renal carcinoma subtypes using multi-tiered network methods

Dana-Farber Cancer Inst, Boston MA

Investigators

Abstract

PROJECT SUMMARY/ABSTRACT Small sample size in Translocation Renal Cell Carcinoma (tRCC), a rare cancer, hampers our ability to correctly diagnose this aggressive subtype and to provide patients with efficient therapies. The most important causes of tRCC misdiagnosis are its heterogenous presentation and overlap with other subtypes such as the major subtype, clear cell RCC (ccRCC), both of which contribute to making tRCC the most aggressive RCC in both primary and metastatic forms. Previous studies have shown that tRCC is initiated by alterations in regulatory elements (i.e., transcriptions factors, TFs) including a translocation at the TFE3 locus with a variable partner gene, which triggers cancer progression. However activity of these TFs has not been systematically identified. The goals of this work are to characterize subtype-specific gene regulatory elements in tRCC by identifying differentially active TFs, determining their metabolic functions, and genetic variants controlling their activity. In Aim 1, Dr. Ben Guebila will develop an optimization method that uses TF binding data and gene expression to estimate context-specific TF activity profiles in tRCC. This method will be based on a constrained optimization approach that uses prior domain knowledge to guide statistical inference and will enable inference using small sample size for rare cancer. Applications of this method using RNA-Seq data in tRCC cell lines will inform activity of TFE3 fusion protein and its implication in oncogenic processes. In Aim 2, Dr. Ben Guebila will build a predictive metabolic network for tRCC cell lines to model various translocation types in TFE3 and inform heterogeneity in tRCC presentation. This model will be then used to identify subtype-specific metabolic pathways activation and find metabolic vulnerabilities that can be exploited as drug targets. In Aim 3, Dr. Ben Guebila will further characterize genetic variants associated to activity of altered TFs in tRCC to determine variable response (including resistance) to recently developed drugs that target these TFs. By developing methods for three network types (transcriptional, metabolic, regulatory), results from this project will allow us to improve diagnosis of tRCC, provide a better molecular definition using multi-omic data, and predict response for newly approved FDA compounds using patient genotype. These research aims are supported by a team of experts in regulatory genomics who will mentor Dr. Ben Guebila’s in five training goals: 1) genomics of kidney cancer subtypes, 2) multiplex network inference, 3) statistical and population genetics, 4) statistical methods for prior knowledge integration, and 5) career development. The institutional environment at Harvard T.H. Chan School of Public Health and Dana-Farber Cancer Institute provide a unique opportunity for developing multi- disciplinary projects at the interface of cancer genetics, network science, and translational medicine. This award will equip Dr. Ben Guebila with the knowledge, skills, and training to pursue impactful and independent research in computational oncology.