Postdoctoral Associate, Imaging Platform
VarChAMP Project: Characterizing functional impacts of coding variation
- Developed machine learning pipeline to detect protein mislocalization and morphological changes across 1600+ coding variants from Cell Painting high-dimensional profiles
- Implemented Bayesian framework to translate functional assay scores into pathogenicity estimates for clinical variant interpretation
OASIS Consortium: Replacing animal testing with high-dimensional profiling
- Developed computational framework for predicting human and rat hepatotoxicity from Cell Painting profiles, laying groundwork for consortium's broader effort integrating multi-modal profiling data to replace animal testing for novel chemicals in medicine and agriculture
VISTA Consortium: Drug repurposing for rare diseases
- Building framework for image-based drug repurposing in rare disease, using optical pooled screening to identify FDA-approved compounds that rescue disease variant-induced protein mislocalization
JUMP Consortium: Benchmarking cell representations
- Co-developing community benchmarks for cell representation learning, defining a compact reference dataset from JUMP, the world's largest public Cell Painting dataset, to enable rapid model training and evaluation
Senior Scientist, Bioinformatics, Translational Discovery, Research and Medicine
- Drug development support: Led bioinformatics initiatives for over 30 projects across all stages of drug development by analyzing multi-modal biological data to identify drug targets, drug mechanisms of action, and clinical biomarkers, accelerate the development processes, and enhance therapeutic precision
- Computational model design: Implemented advanced deep learning models that integrate clinical, genomic, transcriptomic, and imaging data to accurately predict patient drug responses and potential adverse events during immunotherapy treatments
- Innovative workflow development: Developed novel computational workflows for the drug development of mRNA cancer vaccines and cell therapies, including prioritizing new drug targets, selecting candidate molecules, and evaluating cross-reactivity and toxicity using deep learning algorithms
- Bioinformatic tools implementation: Created web tools to generate target-drug-disease networks for new target discoveries and indication expansions, leveraging information from academic publications and in-house biomarker data using GPT
Algorithm Engineer Intern
- In-silico antibody design and optimization: Constructed machine learning frameworks using evolutionary information of B cells and BCR sequences for antibody design and optimization
Graduate Research Assistant
- Genomic resolution prediction: Derived an analytical solution for genomic resolutions in Bulk Segregant Analysis experiments and validated the theoretical results through empirical simulations
- Evolutionary interaction simulation: Implemented and evaluated the evolutionary dynamics between Drosophila and Wolbachia, discovering selection patterns consistent with an arms race model
- Drosophila melanogaster pesticide resistance prediction: Designed a machine learning framework incorporating feature selection and neural networks to predict resistance to different pesticides in Drosophila melanogaster
