I develop computational models in close collaboration with wet laboratories and medical imaging experts, using data-driven approaches to optimize and validate model predictions. My research group focuses on addressing fundamental biological questions and translational challenges across multiple domains, including cell development, regenerative medicine, and disease modeling.

In cell development, we explore the mechanisms controlling neural tube patterning, the role of cell division inheritance in T-cell fate commitment, and the core circuits governing B-cell commitment.

For regenerative medicine, we investigate how to optimize the production of dopaminergic neurons, improve direct conversion of human skin cells to neurons and microglia, and identify gene circuits for reprogramming glial cells into GABAergic neurons. We also aim to increase the yield of PV+ and SST+ neurons and refine the generation of induced pluripotent stem cells using hybrid probabilistic and deep learning models trained on single-cell data.

In the fight against cancer and neurodegenerative diseases, we apply machine learning and advanced imaging techniques (e.g., multispectral and photoacoustic imaging) to assist pathologists in tumor classification, identify cancer driver genes from epigenetic data, and uncover molecular signatures of Parkinson’s disease by combining generative deep models with spatial microspectroscopy.

Our work bridges computational innovation with experimental validation, aiming to advance both fundamental science and clinical applications.