About

I am a Research Fellow at Massachusetts General Hospital and Harvard Medical School, working in Prof. Luca Pinello's group. My work sits at the intersection of artificial intelligence, therapeutic discovery, and computational biology.

I received my PhD in Computer Science from City University of Hong Kong, advised by Prof. Ka-Chun Wong. I was previously a visiting scholar with Prof. Manolis Kellis at MIT and Prof. Marinka Zitnik at Harvard.

My long-term goal is to develop AI systems that make biology more computable, predictable, and actionable. I am particularly interested in generative and foundation models for molecular design and cellular systems, as well as emerging agentic approaches for scientific discovery.

  • Generative therapeutics. Molecular and target-aware design for therapeutic discovery.
  • Computational cell models. Modeling cellular state, regulation, and perturbation from single-cell and multi-omic data.
  • Agentic AI for science. Exploring systems that connect biological hypotheses, data, and design decisions.

Giving an oral talk at ISMB 2026.

TFScope: Learning the DNA-binding specificity of transcription factors from protein sequence will be presented in the RegSys Track.

Our Alzheimer's single-cell epigenomics paper is published in Cell.

Single-cell multiregion epigenomic rewiring in Alzheimer's disease progression and cognitive resilience. Paper

Started as a Research Fellow at MGH / Harvard Medical School.

Joined Prof. Luca Pinello's group to work on generative models for proteins, DNA, and biological element design.

Our single-cell ATAC-seq diffusion model is accepted at NeurIPS 2024.

A versatile informative diffusion model for single-cell ATAC-seq data generation and analysis. Paper

Our 3D molecule generation model is published in Nature Communications.

A dual diffusion model enables 3D binding bioactive molecule generation and lead optimization given target pockets. Paper · Code

Cell 2025 Cell journal cover

Single-cell multiregion epigenomic rewiring in Alzheimer's disease progression and cognitive resilience

Collaborative work on disease progression, cognitive resilience, and single-cell epigenomic regulation.

Overview of the Alzheimer's disease single-cell epigenomics study
Figure 1 from Liu et al., Cell (2025), CC BY-NC-ND 4.0.
NeurIPS 2024 Neural Information Processing Systems logo

A versatile informative diffusion model for single-cell ATAC-seq data generation and analysis

A generative framework for chromatin accessibility data, developed during my visiting scholar work with the Kellis Lab.

Overview of the ATAC-Diff framework
Figure 1 from Huang et al., NeurIPS (2024).
Nature Communications 2024 Nature Communications cover image

A dual diffusion model enables 3D binding bioactive molecule generation and lead optimization

Target-pocket-conditioned molecular generation for structure-aware lead optimization.

Overview of the PMDM target-conditioned molecular generation framework
Figure 1 from Huang et al., Nature Communications (2024), CC BY 4.0.