profile_picture
Zehui Li
PhD, Imperial College London
zl6222@ic.ac.uk

I’m a Ph.D. student studying Machine Learning and Bioinformatics at the Imperial College London advised by Dr. Aaron Zhao and Prof. Guy-Bart Stan. I’m a member of IC-CSynB and DeepWok Lab.

At present, I am funded by BBSRC AI-4-EB(UKRI) for a Ph.D. in Machine Learning applied to Transcriptomics and Genomics sequences. My research focuses on designing deep learning architectures for biological sequence prediction and generation, as well as constructing high-quality open-source models and datasets to bridge the gap between bioinformatics and machine learning.

Prior to this, I worked in Microsoft Shanghai Azure PaaS Team. Before that, I finished my MPhil degree in Advanced Computer Science from Cambridge University and was supervised by Prof. Pietro Lio’ and Prof. Simone Teufel.

Interests

  • Machine learning for transcriptomics and genomics sequences
  • Machine Learning with Graphs
  • Bioinfomatics and Network biology
  • Natural Language Processing
  • Structured Probabilistic Model

Academia

Imperial College London
2023 - present
Ph.D. Bioinformatics and Machine Learning
supervised by Prof. Guy-Bart Stan and Dr. Aaron Zhao
University of Cambridge
2019 - 2020
MPhil Advanced Computer Science
graduated with Distinction, supervised by Prof. Pietro Lio’ and Prof. Simone Teufel
University of Nottingham
2015 - 2019
B.Sc. Computer Science , Minor: Statistics
Graduate with First Class Honours

publications and preprints

Publications and Preprints
Hybrid Graph: A Unified Graph Representation with Datasets and Benchmarks for Complex Graphs, 2023, In Review
Zehui Li* , Xiangyu Zhao* , Mingzhu Shen , Guy-Bart Stan , Pietro Liò , Yiren Zhao
A Hierarchical Genomic Deep Neural Network with 1D Shifted Window Transformer, 2023, ICML 2023 Workshop on Computational Biology
Zehui Li , Akashaditya Das , William A V Beardall , Yiren Zhao , Guy-Bart Stan

Projects

Researh project related to machine learning and bioinformatics
Optimising representation learning of heterogeneous cancer data
An exploration on the optimization routines of SVI for GPs
Adversarial Attack on State-of-the-art Question-Answering Systems