An immune world model for multiscale forecasting and therapeutic hypothesis generation
Abstract
Immune therapies act across cell-intrinsic programs, tissue ecosystems, and patient-specific immune states, yet most predictors address these scales separately. We used a governed evolutionary AI Scientist to construct the Immune World Model, an action-conditioned model that learns how interventions move immune states across cellular, tissue, and individual levels. The Immune World Model--building Scientist searched candidate architectures and workflows, and the resulting world model was frozen before independent confirmation. The frozen model generalized to unseen interventions and biological contexts, recovered intervention-specific cellular programs, integrated cell and tissue information to improve ecosystem and patient-response prediction, and forecast unseen perturbation combinations. Immune World Model--guided analysis then combined measured perturbations with cross-axis inference to nominate IL-36γ plus SIRPα inhibition as a complementary-axis therapeutic hypothesis, whereas a governed self-correction audit rejected every screened cytokine pair. The Immune World Model provides a framework for multiscale immune simulation that connects AI Scientist-driven model construction, intervention forecasting, and the generation of prospectively testable therapeutic hypotheses.
Related
- Multiscale predictive cellular modeling: integrating hypothesis grammars, digital twins, and multi-omics for In silico oncology and precision theranostics
- Drug-centric knowledge graph: representing therapeutic modality, pharmacology and multiscale biology
- Sparse concept attribution for histomorphological hypothesis generation from whole-slide classifiers
- HypoKG: Evidence-Disciplined Biomedical Hypothesis Generation Beyond Endpoint Knowledge
- General Purpose AI for Public Data Analysis and Experimental Hypothesis Generation
BibTeX
@misc{cui2026immune,
title = {{An immune world model for multiscale forecasting and therapeutic hypothesis generation}},
author = {Taoyong Cui and Xi Wang and Zonghang Li and Jinchao Ding and Lingsen You and Yuzhi Xu and Wanghan Xu and Fang Wu and Kejun Ying and Wanli Ouyang and Pheng Ann Heng and Ling Yang and Zhenfei Yin and Yingcheng Wu},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.14709},
eprint = {2609.14709},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.14709},
}
RIS
TY - UNPB
TI - An immune world model for multiscale forecasting and therapeutic hypothesis generation
AU - Taoyong Cui
AU - Xi Wang
AU - Zonghang Li
AU - Jinchao Ding
AU - Lingsen You
AU - Yuzhi Xu
AU - Wanghan Xu
AU - Fang Wu
AU - Kejun Ying
AU - Wanli Ouyang
AU - Pheng Ann Heng
AU - Ling Yang
AU - Zhenfei Yin
AU - Yingcheng Wu
PY - 2026
DA - 2026/09/13
PB - arXiv
DO - 10.48550/arxiv.2609.14709
UR - https://arxiv.org/abs/2609.14709
AB - Immune therapies act across cell-intrinsic programs, tissue ecosystems, and patient-specific immune states, yet most predictors address these scales separately. We used a governed evolutionary AI Scientist to construct the Immune World Model, an action-conditioned model that learns how interventions move immune states across cellular, tissue, and individual levels. The Immune World Model--building Scientist searched candidate architectures and workflows, and the resulting world model was frozen before independent confirmation. The frozen model generalized to unseen interventions and biological contexts, recovered intervention-specific cellular programs, integrated cell and tissue information to improve ecosystem and patient-response prediction, and forecast unseen perturbation combinations. Immune World Model--guided analysis then combined measured perturbations with cross-axis inference to nominate IL-36γ plus SIRPα inhibition as a complementary-axis therapeutic hypothesis, whereas a governed self-correction audit rejected every screened cytokine pair. The Immune World Model provides a framework for multiscale immune simulation that connects AI Scientist-driven model construction, intervention forecasting, and the generation of prospectively testable therapeutic hypotheses.
ER -