About
I am a postdoctoral researcher at the University of Pennsylvania. I completed my PhD in Computer Science in 2026, building foundation models and interpretable ML systems for genomics, RNA biology, and clinical data.
I am looking for research scientist, applied scientist, and ML engineering roles.
- Foundation models and LLMs
- Biomedical and clinical AI
- RNA biology and genomics
- Interpretable multimodal ML
-
PhD in Computer Science, 2026
University of Pennsylvania
-
MS in Computer Science, 2020
Duke University
-
BS in Applied Mathematics, 2018
UC San Diego
Experience
Student Researcher, DeepVariant
2025Google Health
Methylation-aware long-read variant calling and phasing, shipped in DeepVariant 1.9.
Software Engineer
2020 – 2021Eureka Industry Cloud by SAP
Full-stack claims platform.
Software Engineer Intern
2018Teradata
Docker / Kubernetes CI/CD for distributed Python and ML workflows.
Postdoctoral Researcher
2026 –University of Pennsylvania
Foundation models, LLM systems, and interpretable MoE methods for biomedical AI.
PhD, Computer Science
2021 – 2026University of Pennsylvania
G4mer, multimodal Alzheimer’s models, and clinical survival methods.
Research Assistant
2019 – 2020Duke University
TF binding, cardiac variant interpretation, and EHR risk models.
Deep Learning and Computational Biology at UPenn. Outstanding TA Award at Duke.
Selected work
Methylation-aware variant calling in DeepVariant
Long-read variant calling and phasing with epigenetic signals, plus T2T evaluation pipelines. Core methods shipped in the official DeepVariant release.
G4mer, an RNA language model
Transcriptome-wide prediction of RNA G-quadruplexes and disease variants. Validated structurally and functionally; invited talks at ISMB, ASHG, and Google Genomics.
Interpretable multimodal clinical AI
Mixture-of-experts models for Alzheimer’s diagnosis and survival analysis across imaging, EHR, and genomics. ICHI · BCB · iLENS preprint
Publications
- G4mer: an RNA language model for transcriptome-wide identification of G-quadruplexes and disease variants. Nature Communications, 2025. paper
- Interpretable Alzheimer’s diagnosis via multimodal fusion of regional brain experts. IEEE ICHI, 2026. paper
- Expert-driven survival machines. ACM-BCB, 2026. paper
- iLENS: interpretable LLM-guided mixture-of-experts for neuroimaging survival analysis. EMNLP 2026 (submitted). preprint
- A dynamic agentic framework for clinical tasks via specialized expert orchestration. AAAI (in preparation).
- CRL-JEPA: causal representation learning of cell states via world models. ICLR (in preparation).
- Cooperative DNA-binding mechanisms of transcription factors. Nucleic Acids Research, 2023. paper
- GENESIS: gene-specific models for cardiac variants of uncertain significance. Circulation: Arrhythmia and Electrophysiology, 2022. paper
- Mortality prediction and palliative care allocation after hip fracture. JAMDA, 2021. paper
Contact
I am on the job market and would be glad to talk about research, applied ML, and open roles.
The fastest way to reach me is email or LinkedIn.
- farica.zhuang@pennmedicine.upenn.edu
- 3700 Hamilton Walk, Philadelphia, PA 19104