About

Charlie Hou

CS / AI Engineer · Computational Biology · Boston, MA

I'm a software engineer and AI researcher working at the intersection of graph neural networks, single-cell biology, and ecological systems.


My work is grounded in the AI Virtual Cell (AIVC) framework — the vision that multi-scale neural models can represent and simulate biological behavior from protein mutations to tumor immune microenvironments to entire ecosystems.


I approach biology as an engineer: learning on demand, anchored to implementation, building tools that make biological complexity legible to computation.


Beyond the lab, I'm building a smart greenhouse as a first tangible step toward computational ecology, and exploring how traditional herbal knowledge can be formalized through the lens of modern perturbation biology.

Research Focus
  • AI Virtual Cell (AIVC) Framework
  • Graph Neural Networks for Biology
  • Single-cell Transcriptomics
  • Uncertainty Quantification
  • Computational Herbology
Target

Zitnik Lab, Harvard Medical School

Research AI Engineer · 1–3 year horizon

Core Stack
PyTorch Geometric scVI-tools Scanpy / AnnData FastAPI Streamlit Claude API CELLxGENE Census RDKit
Education

Computer Science · AI

Now

What I'm Working On

Research

Pan-cancer extension of TME-VI — zero-shot transfer to NSCLC using CELLxGENE Census (GSE176021).

Building

CellScout multi-agent RAG system for AIVC literature — knowledge graph + cross-paper synthesis.

Maker

Eco-Vault smart greenhouse — Raspberry Pi sensor stack, AnnData time-series pipeline.

Exploring

Computational herbology pipeline — field collection to Virtual Cell perturbation prediction.