Translational Medical Science

Build and test synthetic EHR data models, then compare privacy and utility so you can study rare disease patterns without exposing real records.

Use NLP and fairness scoring to audit FDA AI medical device summaries, extract cohort demographics, and flag gaps in representation.

Build an active-learning drug screening workflow with Bayesian optimization and ChemProp to rank Mpro candidates faster.

Test whether audio tones change headache pain and heart rate variability, and learn how to run a sham-controlled human study with smartphones.

Build a machine learning pipeline to predict KRAS-G12D inhibitors, rank molecules, and interpret atomic features with ADMET filtering.

Build a UV-C safety curve for retainers or mouthguards and practice image-based biofilm analysis with simple, low-cost tools.

Model cafeteria seating and traffic flow with agent-based simulation, then test which low-cost layout changes cut respiratory spread risk.

Build a machine learning model that reads raw PPG signals and tests whether wearables can flag sepsis hours earlier than vital signs.

Build a knowledge graph and test whether it can predict which drug targets move to Phase III, while practicing data cleaning, features, and validation.

Analyze wearable sleep and activity data to cluster recovery patterns and spot risky post-op trajectories with real medical datasets.