Abstract LB342: Learning contextual determinants of cytokine signaling during immunity

Abstract Cytokine action depends on the context of the receiving cell—its type, location, and cytokine milieu. Previous work has mapped the transcriptional response to cytokine treatments across experimental contexts. However, current reference sets do not reflect the full cell-type continuum; they are frequently derived from non-physiological contexts; and they are treated as static entities that ignore covariation between genes and the multitude of cellular co-signals that determine transcription. In this work, we developed computational approaches that capture gene programs driven by cytokine signaling using single-cell gene expression data. Specifically, we used factor analysis algorithms, guided by gene sets of cytokine targets, to decompose single-cell expression matrices into collections of cytokine-induced gene programs. We applied this approach to the recent Immune Dictionary dataset and human COVID-19 atlases for 31 cytokines. We recapitulated known disease-associated cytokine signals and predicted cell-type-specific cytokine responses. Moreover, we uncoupled primary and secondary cytokine signals induced due to in vivo treatment in the Immune Dictionary. Due to the quantitative nature of our inference, we could also measure interactions of different cytokine signals in sub-populations of cells. With this we identified similarly acting co-signals in tumor associated macrophages from pancreatic and lung cancers. We aim to provide a framework for interpreting complex cytokine signals in single-cell data that can be used to dissect tissue biology. Citation Format: Samuel A. Rose, Russell Kunes, Thomas Walle, Suhani Balachandran, Roshan Sharma, Dana Pe'er. Learning contextual determinants of cytokine signaling during immunity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB342.