Connect With Aresty

RAD Collaboratory SURF
Interpretable Multimodal AI Linking Immune Synapse Imaging and Gene Expression for Precision Immunotherapy
Project Summary
"Immune cells eliminate tumors through tightly regulated interactions with cancer cells, yet the molecular and structural mechanisms guiding this process remain poorly understood. When an immune cell contacts a tumor cell, they form an immunological synapse—a dynamic interface where spatial organization and signaling determine whether the tumor is destroyed or evades immune attack. Each participating cell also carries a unique gene expression signature that encodes its functional state and responsiveness. However, imaging and transcriptomic studies have traditionally been performed in isolation, limiting the ability to link molecular regulation with observable cell behavior. This project addresses that gap by developing an AI-driven multimodal framework capable of learning directly from both image and gene expression data. Built upon our Vertical Cell Pairing (VCP) platform, which has generated millions of high-resolution images and videos of immune–tumor synapses over two decades, together with matched gene expression profiles, this project provides an unprecedented opportunity to model immune communication across molecular and spatial dimensions.

We will train large-scale multimodal AI models to learn shared representations between imaging and genomics, revealing how molecular programs control immune cell organization, signaling, and cytotoxic function. By fusing visual and molecular evidence, the framework will uncover interpretable, biologically grounded patterns that explain how genetic activity translates into immune behavior. The integrated model will not only advance fundamental understanding of immune regulation but also enable predictive evaluation of engineered immune therapies, such as CAR-T and CAR-NK cells, whose clinical outcomes remain variable and difficult to forecast. By identifying molecular–morphological signatures linked to therapeutic potency, the system could transform how immune cell products are tested, optimized, and personalized. This research exemplifies how AI-powered multimodal learning can bridge fundamental biology and translational medicine, driving progress in precision immunotherapy and data-driven biomedical discovery."



Sign in to view more information about this project.