QB3-UCSC Graduate Fellowship for Innovators
My research takes a multimodal approach to improving the reproducibility, characterization, and functional relevance of cortical organoid models by leveraging deep learning for both molecular and electrophysiological analysis. Despite their promise for modeling human brain development and neurological disorders, organoids suffer from variability, immature physiology, and limited standardization. My work addresses these challenges by developing computational tools that quantitatively assess organoid identity, maturation, and network behavior across different experimental conditions.
On the transcriptomic side, I developed SIMS, a scalable and reference-based pipeline for automated cell-type annotation in single-cell RNA sequencing data. SIMS enables researchers to objectively assess cellular composition across organoid protocols, detect stress-associated transcriptional states, and benchmark organoids against primary fetal brain datasets. On the functional side, I created HIPPIE, a deep learning framework that extracts electrophysiological signatures from microelectrode array (MEA) recordings, enabling non-destructive evaluation of neuronal network dynamics and maturation.
Beyond neurodevelopment, these approaches offer broadly applicable solutions for other organ systems, disease models, and translational settings, including drug screening, cell therapy quality control, biomarker discovery, and regenerative medicine. Ultimately, my goal is to advance organoids from descriptive models to standardized, quantifiable, and physiologically validated platforms for biological discovery.