About Hao

Live like an anonymous flower.

Welcome to my homepage! I am Hao Xu (@nyxflower), a second-year CS PhD student at the University of California, San Diego, working under the supervision of Prof. Nuno Banderia. Previously, I was a researcher at Queen’s Computational Systems Biology Lab, led by Prof. Laurence Yang. I also worked as a machine learning consultant with Dr. James Yurkovich at Samumed LLC.

I hold an honoured master’s degree in Computer Science from the University of Sheffield, where I completed a dissertation on multirelational link prediction heterogeneous graph under the supervision of Prof. Haiping Lu, earning the the Fretwell-Downing Prize for the best M.SC. Dissertation. Prior to focusing on computer science, I graduated form Central South University and worked as a geotechnical engineer.

Find me at Github, Google Scholar and LinkedIn.

Research Interests

My research is centered on developing trustworthy machine learning and computational methods to address challenges in biomedical science. Specifically, I am interested in using graph neural networks for mining graph-type bio-data, developing point-cloud-based methods for protein function prediction and peptide identification, and exploring the synergy of machine learning and simulation methods in computational biology. Learn more about my research statement and view a list of all projects.

Selected Publications

• GripNet: Graph Information Propagation on Supergraph for Heterogeneous Graphs [Dataset and code]
  Hao Xu, Shengqi Sang, Peizhen Bai, Ruike Li, Laurence Yang, Haiping Lu
  Pattern Recognition, 2023
• PyKale: Knowledge-Aware Machine Learning from Multiple Sources in Python [Code] [PyPI]
  Haiping Lu, Xianyuan Liu, Shuo Zhou, Robert Turner, Peizhen Bai, Raivo Koot, Mustafa Chasmai, Lawrence Schobs, Hao Xu
  Proceedings of the 31st ACM International Conference on Information and Knowledge Management (CIKM), 2022
• Compound Screening with Deep Learning for Neglected Diseases: Leishmaniasis [Code]
  Jonathan Smith, Hao Xu$^*$, Xinran Li, Laurence Yang, Jahir Gutierrez
  Proceedings of the 16th Machine Learning in Computational Biology (MLCB) meeting, 2022
  ($^*$ joint first author)
• A dynamic metabolic map for diabetes
  Jiao Zhao, Hao Xu, Laurence Yang
  Nature Computational Science, 2021
• $\rightarrow$Full list of publications

Selected Open-Source Tools

• PyKale: A Python library in the PyTorch ecosystem for knowledge-aware machine learning focused on multimodal learning and transfer learning for graphs, images and videos, with supporting models on deep learning and dimensionality reduction. [PyPI] [Homepage]

• APRILE: A Python library provided an XAI framework to reveal the mechanisms underlying adverse drug reactions caused by polypharmacy therapy. [PyPI] [Homepage]

• PoSE-Path: A Python command line tool for APRILE. [Homepage]

• img2latex-mathpix: A light tool for converting images to certain LaTeX equation formats and OCR for Windows and MacOS. [Homepage]

Updates

• [06/05/2024] Excited to give a talk on our recent work, “Benchmarking peptide spectral library search”, at the 72th American Society for Mass Spectrometry (ASMS) Conference, Anaheim, California.

• [05/11/2023] Glad to receive the graduate and postdoctoral fellowship from D.E. Shaw Research, and present my previous research on enhancing interpretability of equivariant neural networks for protein 3D structures.

• [11/10/2022] Glad to give another guest lecture on “Neural Networks for Biological Networks: An Introduction” in CHEE 886 at Queen’s University, Kingston, Canada. [Slides]

• [09/22/2022] Lucky to start a new journey at the University of California San Diego as a CS PhD student. I am advised by Prof. Nuno Banderia :>

• [07/14/2022] Presented “Exploring the Mechanisms of Polypharmacy Side Effects via Two-stage Graph Neural Networks” at the 30th Conference on Intelligent System for Molecular Biology (ISMB), Machine Learning in Computational and Systems Biology (MLCSB) COSI, Madison, Wisconsin. [Video]

• [07/13/2022] Presented “Enhancing interpretability of equivariant neural networks for protein structures” at the 30th Conference on Intelligent System for Molecular Biology (ISMB), 3D Special Interest Group (3DSIG) COSI, Madison, Wisconsin. [Poster]

• [11/29/2021] Glad to give a guest lecture on “Neural Networks for Biological Networks: An Introduction” in CHEE 807 at Queen’s University, Kingston, Canada. [Slides]

• [12/14/2019] Presented “Tri-graph Information Propagation for Polypharmacy Side Effect Prediction” at NeurIPS 2019 (Graph Representation Learning Workshop), Vancouver, Canada. [Poster]