Portrait of Jun Ding

Jun Ding

Affiliate Member
Assistant professor, McGill University, Department of Medicine
Research Topics
Computational Biology
Medical Machine Learning
Representation Learning
Website
Google Scholar

Biography

Jun Ding is an assistant professor in the Department of Medicine of the Faculty of Medicine and Health Sciences at McGill University.

Alongside his team, he is dedicated to employing machine learning techniques to decipher the complex dynamics of cells in various diseases, such as developmental disorders, pulmonary diseases and cancers. The diverse and intricate nature of these conditions necessitates innovative approaches, prompting the use of state-of-the-art single-cell technologies to meticulously profile individual cell states. The result is a rich source of data for our machine learning models.

These technologies present unprecedented opportunities to advance understanding, particularly in fields like developmental and cancer biology. However, the challenge is to develop computational models capable of linking this intricate biomedical data to potential discoveries.

Ding’s primary focus lies in the development and refinement of machine learning methodologies, especially probabilistic graphical models, to effectively analyze, model and visualize both single-cell and bulk omics data, often featuring longitudinal or spatial dimensions. The goal is to harness these advanced machine learning techniques to deepen the comprehension of cellular dynamics, and so develop groundbreaking diagnostic and therapeutic strategies that can significantly benefit public health.

Current Students

Vishvak Raghavan
PhD - McGill University
Principal supervisor :
Yue Li
Github

Publications

Molecular pathways of immune checkpoint inhibitor–induced hepatitis.
Erika Bushatsky
Natasha Ryan
Manuel Flores Molina
Steph A. Pang
Judith Lapierre
Madelyn Abraham
Sonia del Rincón
Jun Ding
Marie Hudson
Wilson H. Miller
2573 Background: Immune checkpoint inhibitor (ICI) related hepatitis is a clinically significant immune-related adverse event (irAE) a… (see more)nd a common cause of treatment interruption. It occurs in roughly 5 to 10 percent of patients receiving anti PD-(L)1 monotherapy and in up to one third of those treated with combination ICI therapy. Despite increasing clinical recognition, the molecular mechanisms and predictive factors underlying ICI hepatitis remain poorly defined. The Montreal Immune-Related Adverse Events (MIRAE)-led hepatitis project aims to characterize the immune cell populations and underlying transcriptional programs associated with ICI-hepatitis pathogenesis. Methods: This translational study is conducted within the MIRAE biobank, a prospective multicenter cohort of ICI-treated patients with and without irAEs. The hepatitis cohort includes patients with longitudinal plasma samples collected at baseline, on treatment, and at irAE onset. Ongoing immune profiling efforts include plasma-based cytokine and chemokine analysis, high-throughput plasma proteomics, and single cell RNA sequencing of PBMCs. Preliminary analysis focused on plasma proteomics. Five patients with high-grade ICI-hepatitis and five ICI-treated controls without irAEs were selected and matched by age, sex, and primary tumor. Plasma samples were analyzed using the SomaScan 11K assay to identify differentially expressed proteins and enriched immune pathways. Results: ICI-related hepatitis was clinically severe, requiring systemic corticosteroids in all cases and additional immunosuppressive therapies in most patients. ICI-hepatitis cases showed significantly higher plasma levels of liver injury markers, including ALT and AST, compared with matched controls. Widespread alterations were observed in the circulating proteome, with strong upregulation of liver-enriched proteins and inflammatory mediators. Gene set enrichment analyses revealed enrichment of liver-associated pathways including xenobiotic and bile acid metabolism, as well as IL-12 signaling, interferon-α and γ, neutrophil-associated pathways, and liver-resident macrophage signatures. Pathway analysis of single cell data revealed enhanced cytotoxic activity of CD8 T cells during ICI hepatitis, as exemplified by upregulation of the CTL and IL-6 pathways. Conclusions: ICI-hepatitis was associated with circulating immune signature characterized by liver injury markers, inflammatory mediators, and enrichment of innate immune pathways. These findings provide molecular insight into the immunopathogenesis of ICI hepatitis and inform future biomarker discovery, druggable pathways, and risk stratification.
2026-05-31
Journal of Clinical Oncology (published)
doi.org
GFETM: Genome Foundation-based Embedded Topic Model for scATAC-seq Modeling
Yimin Fan
Yu Li
Jun Ding
Yue Li
Single-cell Assay for Transposase-Accessible Chromatin with sequencing (scATAC-seq) enables investigation of open chromatin landscapes at si… (see more)ngle-cell resolution, but its analysis remains challenging because of sparsity, noise, and dataset-specific peak vocabularies. Genome Foundation Models (GFMs), pre-trained on large DNA sequence corpora, offer a potential source of transferable sequence information for scATAC-seq modeling. We introduce the Genome Foundation Embedded Topic Model (\model{}), an interpretable framework that combines GFMs with the Embedded Topic Model (ETM) for sequence-informed scATAC-seq analysis. By integrating GFM-derived DNA sequence embeddings into a topic-model decoder, \model{} improves clustering quality on standard benchmarks and captures cell-state-specific transcription factor activity through motif scoring and attention-based interpretation.
2026-05-27
FM4LS @ International Conference on Machine Learning (poster)
openreview.net
openreview.net
Sex-specific hormone-sensitive regulatory architecture in adolescence as a scaffold for depression vulnerability
Gladi Thng
Michel Garcia-Miranda
Kailu Song
Anjali Chawla
Reine Khoury
Minh Nguyen
Gabriella Frosi
Matthew Suderman
David Liao
Natalina Salmaso
Tie Yuan Zhang
Pan Wong Tak
Yashar Zeighami
Jun Ding
Corina Nagy
2026-05-20
bioRxiv (preprint)
doi.org
RFGWRK: a hybrid downscaling framework for high-resolution precipitation mapping in geohazard-prone mountainous regions
Simin Zhang
Zeshuang Zheng
Jun Ding
Shengbing Yang
Yuan Zeng
2026-05-18
International Conference on Remote Sensing, Surveying, and Mapping (published)
doi.org
Dissecting and steering cell dynamics using spatially-informed RNA velocity with veloAgent
Vishvak Raghavan
Brent Yoon
Gregory J Fonseca
Yue Li
Jun Ding
RNA velocity enables inference of cell state transitions from single-cell transcriptomics by modeling transcriptional dynamics from spliced … (see more)and unspliced mRNA. However, existing methods overlook spatial context and struggle to scale to large datasets, limiting insights into tissue organization and dynamic processes. We introduce veloAgent, a deep generative and agent-based framework that estimates gene- and cell-specific transcriptional kinetics while integrating spatial information through agent-based simulations of local microenvironments. By leveraging both molecular and spatial cues, veloAgent improves velocity accuracy and achieves sublinear memory scaling, enabling efficient analysis of large and multi-batch spatial datasets. A distinctive feature of veloAgent is its in silico perturbation module, which allows targeted manipulation of spatial velocity vectors to simulate regulatory interventions and predict their impact on cell fate dynamics. These capabilities position veloAgent as a scalable and versatile framework for dissecting spatially resolved cellular dynamics and guiding cell fate manipulation across diverse biological processes.
2026-05-05
Molecular Systems Biology (published)
doi.org
Papillae growth and molecular responses of juvenile sea cucumbers (Apostichopus japonicus) exposed to different light intensities
Weiyan Li
Ziyu Liu
Yajie Deng
Jiaqi Liu
Jinge Yu
Haoran Xiao
Fenglin Tian
Lingshu Han
Chong Zhao
Jun Ding
2026-04-21
Comparative Biochemistry and Physiology Part D: Genomics and Proteomics (published)
doi.org
Primary large-cell neuroendocrine carcinoma of the prostate and its nursing care: A systematic review
Mingli Wang
Xuemei Zhang
Lifang Pan
Jun Ding
2026-03-31
Asian Journal of Surgery (published)
doi.org
1113 GPR124 Alleviates Blood-Brain Barrier Disruption Through Improving the Function of Microvascular Endothelial After Traumatic Brain Injury
Chen Wang
Jun Ding
Hengli Tian
Hao Chen
Xinyu Niu
2026-03-25
Neurosurgery (published)
doi.org
Summarize Before You Speak with ARACH: A Training-Free Inference-Time Plug-In for Enhancing LLMs via Global Attention Reallocation
Jingtao Wang
Yucong Wang
Jun Ding
Rui Cai
Xun Wang
Large language models (LLMs) achieve remarkable performance, yet further gains often require costly training. This has motivated growing int… (see more)erest in post-training techniques-especially training-free approaches that improve models at inference time without updating weights. Most training-free methods treat the model as a black box and improve outputs via input/output-level interventions, such as prompt design and test-time scaling through repeated sampling, reranking/verification, or search. In contrast, they rarely offer a plug-and-play mechanism to intervene in a model's internal computation. We propose ARACH(Attention Reallocation via an Adaptive Context Hub), a training-free inference-time plug-in that augments LLMs with an adaptive context hub to aggregate context and reallocate attention. Extensive experiments across multiple language modeling tasks show consistent improvements with modest inference overhead and no parameter updates. Attention analyses further suggest that ARACH mitigates the attention sink phenomenon. These results indicate that engineering a model's internal computation offers a distinct inference-time strategy, fundamentally different from both prompt-based test-time methods and training-based post-training approaches.
2026-03-09
arXiv (preprint)
doi.org
arxiv.org
Research on Hybrid Deep Learning Prediction Method for Midship Vertical Bending Moments
Jun Ding
Peiqiao Zhu
zhang Zhu
Yiming Qiang
2026-01-31
Oceans (published)
doi.org
Dissecting and steering cell dynamics using spatially-informed RNA velocity with veloAgent
Brent Yoon
Vishvak Raghavan
Jun Ding
RNA velocity enables inference of cell state transitions from single-cell transcriptomics by modeling transcriptional dynamics from spliced … (see more)and unspliced mRNA. However, existing methods overlook spatial context and struggle to scale to large datasets, limiting insights into tissue organization and dynamic processes. We introduce veloAgent, a deep generative and agent-based framework that estimates gene- and cell-specific transcriptional kinetics while integrating spatial information through agent-based simulations of local microenvironments. By leveraging both molecular and spatial cues, veloAgent improves velocity accuracy and achieves sublinear memory scaling, enabling efficient analysis of large and multi-batch spatial datasets. A distinctive feature of veloAgent is its in silico perturbation module, which allows targeted manipulation of spatial velocity vectors to simulate regulatory interventions and predict their impact on cell fate dynamics. These capabilities position veloAgent as a scalable and versatile framework for dissecting spatially resolved cellular dynamics and guiding cell fate manipulation across diverse biological processes.
2026-01-08
bioRxiv (preprint)
doi.org
SIDISH integrates single-cell and bulk transcriptomics to identify high-risk cells and guide precision therapeutics through in silico perturbation
Yasmin Jolasun
Kailu Song
Yumin Zheng
Jingtao Wang
Gregory Fonseca
David H. Eidelman
Jun Ding
Single-cell RNA sequencing (scRNA-seq) provides high-resolution insights into cellular heterogeneity but remains costly, restricting its use… (see more) to small cohorts that often lack comprehensive clinical data, reducing translational relevance. In contrast, bulk RNA sequencing is scalable and cost-effective but obscures critical single-cell insights. We introduce SIDISH, a neural network framework that integrates the granularity of scRNA-seq with the scalability of bulk RNA-seq. Using a variational autoencoder, deep Cox regression, and transfer learning, SIDISH identifies high-risk cell populations while enabling robust clinical predictions from large-cohort data. Its in silico perturbation module identifies therapeutic targets by simulating interventions that reduce high-risk cells associated with adverse outcomes. SIDISH also generalizes to spatial transcriptomics, identifying high-risk cells and mapping them within their native tissue microenvironment. Applied across diverse diseases, SIDISH establishes the link between cellular dynamics and clinical phenotypes, facilitating biomarker discovery and precision medicine. By unifying single-cell insights with large-scale clinical data, SIDISH advances computational tools for disease risk assessment and therapeutic prioritization, offering an integrative and scalable approach to precision medicine. SIDISH integrates single-cell and bulk RNA sequencing data using deep learning to identify high-risk cell populations and prognostic biomarkers, enabling in silico perturbations that could guide precision therapeutics and advance personalized medicine.
2025-12-09
Nature Communications (published)
doi.org