Awarded in 2022
Home Department: Computer Science
Faculty Advisors: Anshul Kundaje (Genetics and Computer Science) and Maya M. Kasowski (Medicine and Pathology)

Research Title: Base-resolution Deep Learning Models of Bulk and Single Cell Multi-omic data to Decipher the Regulatory Basis of Thyroid Cancer

Research Description: Gene expression regulation and dynamics associated with thyroid cancer development remain unexplored. In collaboration with the Kundaje and Kasowski labs, Anusri proposes computational deep learning models to accurately predict regulatory (chromatin accessibility) profiles from multi-omic high-resolution sequence data. Anusri will train models on bulk and single cell multi-omic data collected from diverse cohorts of healthy thyroid, neoplasm, and metastases. Model interpretation will decipher the context-specific regulatory syntax of dynamics across different axes of variation, with a specific focus on developing hypotheses regarding regulatory wiring of the iodine channel in normal thyroid and cancer. Anusri’s successful training of deep learning models along with targeted CRISPR experimentation to validate hypotheses will aid in the efforts to understand the regulatory basis of thyroid cancer.