AI for Immunology: Toward a Virtual Immune System

February 27, 2026
10:00 am to 11:00 am

Event sponsored by

Biostatistics and Bioinformatics
BERD Core
Center for AIDS Research (CFAR)
Department of Surgery
School of Medicine (SOM)

Contact

Adkins, Judy

Share

Chen Jinmian, PhD

Speaker

Chen Jinmian, Associate Professor, Centre for Computational Biology, Duke-NUS

Abstract: The immune system is a complex, adaptive network spanning molecules, cells, tissues, and organs. With the advent of high-dimensional single-cell and spatial omics, immunology is entering a data-rich era and that urgently demands integrative and interpretable computational frameworks. In this talk, I will discuss how artificial intelligence (AI) can transform immunological discovery and translation. I will highlight our efforts to build virtual immunity, an AI-driven framework that models the immune system at cellular, tissue, and systemic levels. This includes deep-learning methods for large-scale single-cell integration (DISCO), spatial organization (GraphST, SpatialGlue, STAMP), and digital twin modeling (ImmunoTwin) that predicts immune aging and responses across contexts. Together, these approaches aim to construct a human immune atlas and an interpretable, data-driven "virtual immune system" to accelerate precision immunology and therapeutic innovation. The immune system is a complex, adaptive network spanning molecules, cells, tissues, and organs. With the advent of high-dimensional single-cell and spatial omics, immunology is entering a data-rich era and that urgently demands integrative and interpretable computational frameworks.

In this talk, I will discuss how artificial intelligence (AI) can transform immunological discovery and translation. I will highlight our efforts to build virtual immunity, an AI-driven framework that models the immune system at cellular, tissue, and systemic levels. This includes deep-learning methods for large-scale single-cell integration (DISCO), spatial organization (GraphST, SpatialGlue, STAMP), and digital twin modeling (ImmunoTwin) that predicts immune aging and responses across contexts. Together, these approaches aim to construct a human immune atlas and an interpretable, data-driven "virtual immune system" to accelerate precision immunology and therapeutic innovation.

Zoom: https://duke.zoom.us/j/93703663288?pwd=9zRBNe7RMcxCGAKGUfTZJx0MXuMaAh.1
Meeting Id: 937 0366 3288
Passcode: 103794

Event Series

Seminar Series

Share