Third-year Ph.D. student · University of California, Irvine
Yaqi Hu
Yaqi Hu is a third-year Ph.D. student at the University of California, Irvine. His research focuses on computational methods for bioinformatics, particularly pathology image analysis and spatial transcriptomics. His broader interests include computational biology, machine learning and deep learning.
NeurIPS 2026sMMC-22M: A Context-Aware Dataset and Benchmark for Single-Cell Spatial Transcriptomics was accepted to the main track.
KDD 2026MUGO: Differentiable Combinatorial Optimization for Causal Variant Discovery in the Non-coding Genome was published in the KDD 2026 proceedings.
CIKM 2026STMoE: Multi-Scale Mixture-of-Experts for Single-Cell Gene Expression Prediction from Histology was accepted.
JudgeJudged oral and poster presentations at the 2026 UC Irvine Undergraduate Research Symposium.
JudgeJudged student projects in AI, machine learning and software engineering at IrvineHacks.
ACML 2024Presented our lab’s paper Understanding Transcriptional Regulatory Redundancy by Learnable Global Subset Perturbations in Hanoi. Best Student Paper Award.
sMMC-22M: A Context-Aware Dataset and Benchmark for Single-Cell Spatial Transcriptomics
Author list coming with the camera-ready version
A single-cell spatial transcriptomics dataset and benchmark that keeps each cell’s tissue context.
Paper and code after camera-ready
KDD 2026Jeju, Korea
MUGO: Differentiable Combinatorial Optimization for Causal Variant Discovery in the Non-coding Genome
S. D. Sun, J. Liu, P. Xu, Y. Hu, M. J. Zhang, J. Zhang
Turns variant discovery into gradient-based optimization over DNA sequence, so a foundation model can search single- and multi-variant edits efficiently.
@inproceedings{sun2026mugo,
title = {MUGO: Differentiable Combinatorial Optimization for Causal Variant Discovery in the Non-coding Genome},
author = {Sun, Simon Dongbo and Liu, Junhao and Xu, Pengcheng and Hu, Yaqi and Zhang, Martin Jinye and Zhang, Jing},
booktitle = {Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2},
pages = {12042--12053},
year = {2026},
doi = {10.1145/3770855.3818971}
}
Main figure coming soon
CIKM 2026Rome, Italy
STMoE: Multi-Scale Mixture-of-Experts for Single-Cell Gene Expression Prediction from Histology
Author list coming with the camera-ready version
Predicts single-cell gene expression from histology images, with experts that look at the tissue at several scales.
