Xiao Wang

Xiao Wang

Postdoc@University of Washington
CS Ph.D.@Purdue University
E-mail: wang3702 [at] uw [dot] edu


About Me

I am now a postdoc of Noble Research Lab and Sheng Wang's Lab, under the supervision of Prof. William Stafford Noble and Prof. Sheng Wang. Prior to that, I obtained a Computer Science Ph.D. degree from Department of Computer Science, Purdue University, advised by Prof. Daisuke Kihara. My research interests lie in computational biology, self-supervised learning, as well as all other intelligent systems.

Starting from 2018, I mainly worked with Prof. Daisuke Kihara on the macromolecular structure modeling, prediction and evaluation. In summer 2019, I did an internship in Futurewei AI Lab supervised by Dr. Lin Chen, Prof. Guo-Jun Qi and Prof. Jiebo Luo. In summer 2020, I interned in JD AI Research supervised by Dr. Jingen Liu and Prof. Jiebo Luo. In summer 2021, I did internship in Facebook AI Research supervised by Dr. Xinlei Chen, Dr. Yuandong Tian and Haoqi Fan. During internships, my research focus is self-supervised learning(SSL).

Before that, I graduated with a bachelor's degree in computer science from Xi'an Jiaotong University, Xi'an, China. During my undergraduate, I mainly worked on intelligent transportation systems under the supervision of Prof. Li Li from Tsinghua University and Prof.Fei-Yue Wang from State Key Laboratory of Management and Control for Complex Systems of Chinese Academy of Sciences. Also, I did a summer intern at Purdue in 2017, working on protein model evaluation supervised by Prof. Daisuke Kihara.

Fellowship&Awards

  • Jan 2024 - Dec 2024: UW Data Science Postdoctoral Fellow
  • July 2022 - June 2023: NSF MolSSI Fellowship with $80,000 for stipend, tuition and fees.
  • Aug 2018 - July 2019: Chiang Chen Overseas Fellowship with $50,000 for tuition and fees.
  • Sep 2017 - July 2018: HIWIN Outstanding Student Scholarship with ¥10,000 for tuition.
  • Dec 2017: Top 10 Undergraduate of Xi'an Jiaotong University.
  • Sep 2016 - July 2017: National Scholarship with ¥8,000 for tuition.
  • Recent News

  • Jul 2024: DiffModeler got accepted in Nature Methods.
  • Jan 2024: DeepMainMast is selected as the cover of Volume 21 issue 1 of Nature Methods.
  • Sep 2023: DeepMainMast got accepted in Nature Methods.
  • May 2023: CryoREAD got accepted in Nature Methods and selected for research briefing.
  • March 2023: DAQ-Database got accepted in Nature Methods.
  • March 2023: CaCo got accepted in IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE T-PAMI).
  • March 2023: CoSeg got accepted in IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS).
  • July 2022: My old paper GRU-CF was nominated for George N. Saridis Best Transactions Paper Award.
  • June 2022: DAQ got accepted in Nature Methods.
  • May 2022: I got NSF MolSSI Fellowship by MolSSI of National Science Foundation (NSF).
  • Selected Publications

    Computational Biology

    DiffModeler: Large Macromolecular Structure Modeling for Cryo-EM Maps Using Diffusion Model
    Xiao Wang, Han Zhu, Genki Terash, Manav Taluja, Daisuke Kihara
    Nature Methods, 2024
    Paper Code Server
    De Novo Structure Modeling for Nucleic Acids in cryo-EM Maps Using Deep Learning
    Xiao Wang, Genki Terash, Daisuke Kihara
    Nature Methods, 2023
    Paper Code Colab Server
    Selected for research briefing CryoREAD provides fully automated DNA–RNA structure modeling for cryo-EM maps in Nature Methods.
    DeepMainmast: integrated protocol of protein structure modeling for cryo-EM with deep learning and structure prediction
    Genki Terash, Xiao Wang, Devashish Prasad, Tsukasa Nakamura, Daisuke Kihara
    Nature Methods, 2023
    Paper Code Colab Server
    Selected as the cover of Volume 21 issue 1 of Nature Methods
    DAQ-Score Database: Assessment of Map-Model Compatibility for Protein Structure Models from Cryo-EM Maps
    Tsukasa Nakamura, Xiao Wang, Genki Terash, Daisuke Kihara
    Nature Methods, 2023
    Paper Database Server
    Residue-Wise Local Quality Estimation for Protein Models from Cryo-EM Maps
    Genki Terash*, Xiao Wang*, Sai Raghavendra Maddhuri Venkata Subramaniya, John J. G. Tesmer, Daisuke Kihara
    Nature Methods, 2022
    Paper Code Colab Server
    Protein Model Refinement for Cryo-EM Maps Using DAQ score
    Genki Terashi, Xiao Wang, Daisuke Kihara
    Acta Crystallographica Section D: Structural Biology, 2022
    Paper Code Colab Server
    Detecting Protein and DNA/RNA Structures in Cryo-EM Maps of Intermediate Resolution Using Deep Learning
    Xiao Wang, Eman Alnabati, Tunde W Aderinwale, Sai Raghavendra Maddhuri, Genki Terashi, Daisuke Kihara
    Nature Communications, 2021
    Paper Code Code_Ocean Colab Server
    Protein Docking Model Evaluation by Graph Neural Networks
    Xiao Wang, Sean T Flannery, Daisuke Kihara
    Frontiers in Molecular Biosciences, 2021
    Paper code
    Protein Docking Model Evaluation by 3D Deep Convolutional Neural Network
    Xiao Wang, Genki Terashi, Charles W Christoffer, Mengmeng Zhu, Daisuke Kihara
    Bioinformatics , 2019
    Paper code Server

    Self-Supervised Learning

    On the Importance of Asymmetry for Siamese Representation Learning
    Xiao Wang*, Haoqi Fan*, Yuandong Tian, Daisuke Kihara, Xinlei Chen
    Conference on Computer Vision and Pattern Recognition (CVPR), 2022
    Paper Code
    CaCo: Both Positive and Negative Samples are Directly Learnable via Cooperative-adversarial Contrastive Learning
    Xiao Wang, Yuhang Huang, Dan Zeng, Guo-Jun Qi
    IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE T-PAMI), 2023
    Paper Code
    Adco: Adversarial contrast for efficient learning of unsupervised representations from self-trained negative adversaries
    Xiao Wang*, Qianjiang Hu*, Wei Hu, Guo-Jun Qi
    Conference on Computer Vision and Pattern Recognition (CVPR), 2021
    Paper Code
    Contrastive Learning with Stronger Augmentation
    Xiao Wang, Guo-Jun Qi
    IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE T-PAMI), 2022
    Paper Code
    EnAET: Self-Trained Ensemble AutoEncoding Transformations for Semi-Supervised Learning
    Xiao Wang, Daisuke Kihara, Jiebo Luo, Guo-Jun Qi
    IEEE Transactions on Image Processing (IEEE TIP) , 2020
    Paper code
    Learning generalized transformation equivariant representations via autoencoding transformations
    Guo-Jun Qi, Liheng Zhang, Feng Lin, Xiao Wang
    IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE T-PAMI), 2020
    Paper code
    CoSeg: Cognitively Inspired Unsupervised Generic Event Segmentation
    Xiao Wang, Jingen Liu, Tao Mei, Jiebo Luo
    IEEE Transactions on Neural Networks AND Learning Systems (IEEE TNNLS), 2022
    Paper code

    Intelligent Transportation

    Capturing Car-Following Behaviors by Deep Learning
    Xiao Wang, Rui Jiang, Li-Li, Yilun Lin, Xinhu Zheng, Fei-Yue Wang
    IEEE Transactions on Intelligent Transportation Systems (IEEE-T-ITS), 2017
    Paper
    Nominated for George N. Saridis Best Transactions Paper Award.
    Long memory is important: A test study on deep-learning based car-following model
    Xiao Wang, Rui Jiang,Li-Li, Yilun Lin, Fei-Yue Wang
    Physica A: Statistical Mechanics and its Applications, 2019
    Paper

    Open Source Projects

  • DiffModeler:DiffModeler is a computational tool using diffusion model to automatically build full protein complex structure by taking native/AF2 single-chain structure as input.  GitHub stars
  • DeepMainMast: DeepMainMast is a computational tool using deep learning to automatically build full protein complex structure from cryo-EM map.  GitHub stars
  • CryoREAD: Cryo_READ is a computational tool using deep learning to automatically build full DNA/RNA atomic structure from cryo-EM map.  GitHub stars
  • DAQ_Refine: DAQ_Refine is a protein structure refinement tool by DAQ-score and ColabFold.  GitHub stars
  • Asym Siam: Asym-Siam experimentally verified the importance of asymmetry for Siamese Representation Learning with obvious improvement.  GitHub stars
  • CaCo: CaCo is a state-of-the-art cooperative-adversarial contrastive learning method where both positive and negative samples are directly learnable.  GitHub stars
  • DAQ: DAQ is a software accesses the quality of protein models built from cryo-Electron Microscopy (EM) maps.  GitHub stars
  • OC_Finder: OC_Finder is a computational tool using deep learning for fully automated osteoclast segmentation, classification, and counting. GitHub stars
  • CoSeg: CoSeg is a self-supervised learning-based event boundary detection method. GitHub stars
  • CLSA: CLSA is a general contrastive learning framework by introducing the information from stronger augmentation. GitHub stars
  • AdCo: AdCo is an algorithm for effective self-supervised learning through adversarial training of negative examples. GitHub stars
  • GNN DOVE: GNN_DOVE is a software can evaluate the quality of protein-docking models using graph neural networks by reformulating protein structures as graphs. GitHub stars
  • Attention_AD: Attention_AD is a software that can distinguish active and inactive peptides for gene expression using Long Short Term Memory (LSTM). GitHub stars
  • Emap2sec+: Emap2sec+ is a software detects local structure information of proteins and DNA/RNA in cryo-EM maps using deep learning. GitHub stars
  • EnAET: EnAET is a software that benefits semi-supervised learning via self-trained ensemble auto-encoding transformations. GitHub stars GitHub stars
  • DOVE: DOVE is a software can evaluate the quality of protein-docking models using 3D neural networks. GitHub stars