Managing a team of scientists and leading computational strategy for 12+ external collaborations. Driving cross-org adoption of LLM-based workflows for biological data interpretation and research automation.
Led analysis for 13 conference posters in 16 months (CSHL, AACR, SITC, SFN, Gordon Research); 2 high-impact papers in submission with patent filings on platform technology.
Developed production-grade analytical software enabling end-to-end analysis and interpretation of multi-modal single-cell data; established documentation, vignettes, and CI/CD pipelines for reproducible workflows across R&D, assay development, and external collaborations.
Led evaluation of DINOv2 computer vision for morphology-transcriptome linking at single-cell resolution; validated biological interpretability for phenotype prediction.
Developed computational screening pipeline integrating public databases with in-house scRNA-seq to identify 14 essential gene targets; findings enabled genome engineering of inducible safety switches for selective elimination of undifferentiated cells in iPSC therapeutics. Work presented at Keystone Symposium 2024.
Built scalable QC pipelines and interactive dashboards for multi-omics analysis; established release assay workflows for beta islet cell therapy manufacturing supporting the SC451 Type 1 diabetes program through IND-enabling stages.
Robustly characterized GS94 safe harbor locus by surveying 300,000 genomic sites computationally, prioritizing 39 candidates for experimental validation; discovery directly enabled AB-1015 CAR-T program for ovarian cancer, now in Phase 1 clinical trials.
Published in SITC 2022: GS94 Safe Harbor Discovery.
Built multi-modal single-cell atlas of gynecologic tumors: 1.2M+ cells with integrated transcriptomics, proteomics (130+ antibody CITE-seq), and chromatin accessibility across 45 donors. Developed scalable visualization tools enabling cross-functional teams to explore and query the atlas interactively.
Benchmarked ML models for genome-wide CRISPR knock-in efficiency prediction; performed comprehensive off-target validation using complementary methods (iGUIDE, SITE-seq, rhAmpSeq).
Integrated multi-omics data (ATAC-seq, RNA-seq, scRNA-seq) to characterize tumor cell plasticity in metastatic colorectal cancer. Identified NF-kB, Hippo, and inflammatory pathways induced upon cancer stem cell depletion — mechanisms informing strategies to overcome treatment resistance.
Advised by Robert Piskol (Sr. Director, Oncology Bioinformatics).
Established single-cell genomics capability at UCLA — built one of the first Drop-Seq instruments on the West Coast, enabling large-scale profiling across multiple biological systems.
Profiled >200,000 single cells to characterize iPSC reprogramming dynamics and identify conserved gene network silencing mechanisms governing cell fate transitions. Developed a computational method using nearest-neighbor regression to predict cell type-specific chromatin features from gene expression alone, eliminating the need for costly epigenetic assays.
Co-advised by Kathrin Plath (Stem Cell Biology) and Jason Ernst (Computational Biology). 16 publications in Cell, Nature, and related journals spanning stem cell reprogramming, cancer biology, and cell therapy.