Hi, I'm Fang WU!

Welcome to my personal web page! I am a research engineer at Tsinghua University advised by Jinbo Xu. My research focuses on deep learning algorithms for scientific problems — in particular, 3D geometric networks, generative AI, domain adaptation, and other applications in chemistry and structural biology. I obtained my Master's degree at Columbia University, advised by Dragomir Radev. It is a profound loss for me to lose Prof. Dragomir on March 29, 2023 (in memoriam).

Email: fw2359@columbia.edu
Address: Haidian District, Beijing, China
Last update time: 2023.10.25

News and Highlights

   [2024/01] One paper is accepted by ICLR 2024.
   [2023/10] One paper is accepted by Nature Communications.
   [2023/09] One paper is accepted by NeurIPS 2023.
   [2023/06] One paper is accepted by Communications Biology.
   [2023/04] Two papers are accepted by ICML 2023.
   [2023/02] One work on molecular domain adaptation is accepted by Cell Patterns.
   [2023/01] A talk on M2D2 invited by MILA and Valence.
   [2022/12] Two papers are accepted by AAAI 2023.
   [2022/10] One work on 3D pretraining is accepted by Advanced Science.

Research Summary

* represents equal contribution and co-first authorship.

Graph Neural Networks (GNNs)

   Rethinking Explaining Graph Neural Networks via Non-parametric Subgraph Matching
   Fang Wu, Siyuan Li, Dragomir Radev, Stan Z. Li
   ICML 2023
   [Paper]    [Code]

   Discovering and Explaining the Representation Bottleneck of Graph Neural Networks from Multi-order Interactions
   Fang Wu*, Siyuan Li*, Dragomir Radev, Stan Z. Li
   Under Review
   [Paper]    [Code]

3D Geometric Deep Learning for Molecules, Proteins, and Materials

   Molformer: Motif-based Transformer on 3D Heterogeneous Molecular Graphs.
   Fang Wu, Dragomir Radev, Stan Z. Li
   AAAI 2023
   [Paper]    [Code]

   Integration of pre-trained protein language models into geometric deep learning networks.
   Fang Wu, Liong Wu, Dragomir Radev, Jinbo Xu, Stan Z. Li
   Communications Biology
   [Paper]    [Code]

   Direct Prediction of Gas Adsorption via Spatial Atom Interaction Learning.
   Jiyu Cui*, Fang Wu*, Wen Zhang*, Lifeng Yang*, Jianbo Hu, Yin Fang, Peng Ye, Qiang Zhang, Xian Suo, Yiming Mo, Xili Cui, Huajun Chen, Huabin Xing
   Nature Communications
[Paper] [Code]

Pretraining & Semi-supervised Learning

   InstructBio: A Large-scale Semi-supervised Learning Paradigm.
   Fang Wu, Huiling Qin, Siyuan Li, Stan Z. Li, Xianyuan Zhan, Jinbo Xu, Stan Z. Li
   Under Review
   [Paper]    [Code]

   A Hierarchical Training Paradigm for Antibody Structure-sequence Co-design
   Fang Wu, Stan Z. Li
   NeurIPS 2023, MLHC 2023
   [Paper]

Domain Adaptation in Biochemistry

   Metric Learning-enhanced Optimal Transport for Biochemical Regression Domain Adaptation
   Fang Wu*, Nicolas Courty*, Shuting Jin*, Stan Z. Li
   Patterns
   [Paper]    [Code]

Molecular Dynamics (MD) Simulations

   Pre-training of Equivariant Graph Matching Networks with Conformation Flexibility for Drug Binding
   Fang Wu*, Shuting Jin*, Yinghui Jiang*, Xurui Jin, Bowen Tang, Zhangming Niu, Qiang Zhang, Xiangxiang Zeng, Stan Z. Li
   Advanced Science
   [Paper]    [Code]

   DiffMD: A Geometric Diffusion Model for Molecular Dynamics Simulations
   Fang Wu, Stan Z. Li
   AAAI 2023 (Oral)
   [Paper]

Others (e.g., NLP, CV)

   InsertGNN: Can Graph Neural Networks Outperform Humans in TOEFL Sentence Insertion Problem?
   Fang Wu, Stan Z. Li
   arXiv 2022
   [Paper]    [Data]

   Architecture-Agnostic Masked Image Modeling: From ViT back to CNN
   Siyuan Li*, Di Wu*, Fang Wu, Zelin Zang, Kai Wang, Lei Shang, Baigui Sun, Hao Li, Stan Z. Li
   ICML 2023
   [Paper]    [Code]

   SemiReward: A General Reward Model for Semi-supervised Learning
   Siyuan Li*, Weiyang Jin*, Zedong Wang*, Fang Wu,, Zicheng Liu, Cheng Tan, Stan Z. Li
   ICLR 2024
   [Paper]

Education

   Columbia University, 2019-2021
   • Master of Science
   • GPA: 3.51/4.0

   Central University of Economics and Finance, 2015-2019
   • Bachelor of Science
   • GPA: 3.85/4.0

Research Experience

Before joining Tsinghua University, I feel fortunate to be a research assistant advised by Stan Z. Li at Westlake University, and recieved guidance as a visiting student from Huajun Chen, Xiang Bai and Danny Lan.

   Research Engineer (2022.08-Now)
   • Tsinghua University
   • Advised by Jinbo Xu

   Research Assistant (2021.11-2022.07)
   • Westlake University
   • Advised by Stan Z. Li

   Visiting Student (2021.03-2021.10)
   • Zhejiang University
   • Hosted by Huajun Chen

Acknowledgement

My study cannot be possible without the support from my awesome friends, mentors, and collaborators! Check out some of them:

Prof. Dragomir Radev, Dr. Xiangru Tang at Yale University. R.I.P. to Dr. Dragomir.

Prof. Stan Z. Li, Prof. Danny Lan, Dr. Siyuan Li, Dr. Lirong Wu, Dr. Haitao Lin at Westlake University.

Prof. Jinbo Xu, Prof. Xianyun Zhan, Dr. Shikun Feng at Tsinghua University.

Prof. Huajun Chen, Prof. Qiang Zhang, Prof. Huabing Xing, Dr. Jiyu Cui at Zhejiang University.

Prof. Wenbing Huang at Renmin University.

Dr. Xiangxiang Zeng at Hunan Technology University.

Prof. Shuting Jin at Wuhan Technology University.

Prof. Nicolas Courty at University Bretagne Sud.

Prof. Buyong Ma, Prof. Shuangjia Zheng , Prof. Junchi Yan at Shanghai Jiaotong University

Prof. Xiang Bai at Huazhong University of Science and Technology.

Aside from university collaborations, I also collaborated with many industrial companies as well as non-profit organizations, including MindrankAI, MoleculeMind, and Biomap

Dr. Xurui Jin, Dr. Yinghui Jiang, Dr. Zhangming Niu at MindrankAI.

Prof. Le Song, Dr. Taifeng Wang, Dr. Linlin Chao at BioMap.

Dr. Guojiang Zhao at DeepPotential.