Integrative mapping of preexisting influenza immune landscapes predicts vaccine response
Stephanie Hao, … , Thushan I. de Silva, Adriana Tomic
Stephanie Hao, … , Thushan I. de Silva, Adriana Tomic
Published July 15, 2025
Citation Information: J Clin Invest. 2025;135(18):e189300. https://doi.org/10.1172/JCI189300.
View: Text | PDF
Clinical Research and Public Health Clinical Research Immunology Virology

Integrative mapping of preexisting influenza immune landscapes predicts vaccine response

Abstract

BACKGROUND Predicting individual vaccine responses is a substantial public health challenge. We developed Immunaut, an open-source, data-driven framework for systems vaccinologists to analyze and predict immunological outcomes across diverse vaccination settings, beyond traditional assessments.METHODS Using a comprehensive live attenuated influenza vaccine (LAIV) dataset from 244 Gambian children, Immunaut integrated prevaccination and postvaccination humoral, mucosal, cellular, and transcriptomic data. Through advanced modeling, our framework provided a holistic, systems-level view of LAIV-induced immunity.RESULTS The analysis identified 3 distinct immunophenotypic profiles driven by baseline immunity: (a) CD8+ T cell responders with strong preexisting immunity boosting memory T cell responses; (b) mucosal responders with prior influenza A virus immunity developing robust mucosal IgA and subsequent influenza B virus seroconversion; and (c) systemic, broad influenza A virus responders starting from immune naivety who mounted broad systemic antibody responses. Pathway analysis revealed how preexisting immune landscapes and baseline features, such as mucosal preparedness and cellular support, quantitatively dictate vaccine outcomes.CONCLUSION Our findings emphasize the power of integrative, predictive frameworks for advancing precision vaccinology. The Immunaut framework is a valuable resource for deciphering vaccine response heterogeneity and can be applied to optimize immunization strategies across diverse populations and vaccine platforms.FUNDING Wellcome Trust (110058/Z/15/Z); Bill & Melinda Gates Foundation (INV-004222); HIC-Vac Consortium; NIAID (R21 AI151917); NIAID CEIRR Network (75N93021C00045).

Authors

Stephanie Hao, Ivan Tomic, Benjamin B. Lindsey, Ya Jankey Jagne, Katja Hoschler, Adam Meijer, Juan Manuel Carreño Quiroz, Philip Meade, Kaori Sano, Chikondi Peno, André G. Costa-Martins, Debby Bogaert, Beate Kampmann, Helder Nakaya, Florian Krammer, Thushan I. de Silva, Adriana Tomic

×

Figure 3

Automated machine learning framework for mapping and predicting LAIV immunogenicity response phenotypes.

Options: View larger image (or click on image) Download as PowerPoint
Automated machine learning framework for mapping and predicting LAIV imm...
(A) Overview of the automated machine learning framework developed to predict LAIV response phenotypes using baseline immune data from mucosal and blood samples, capturing multidimensional immune parameters such as transcriptomics, antibody titers, bacterial load, flu-specific T cell responses, and comprehensive immunophenotyping. (B) Step 1, balanced data partitioning: the dataset is split into training (80%) and testing (20%) sets, ensuring proportional representation of each immunophenotypic group (CD8+ T cell; mucosal; and systemic, broad influenza A responders) to maintain predictive accuracy across classes. Step 2, model optimization cycle: 10-fold cross-validation and hyperparameter tuning are applied across 141 machine learning models, each iteratively trained and validated to identify the best predictors of vaccine response. Step 3, model evaluation and scoring: predictive performance metrics, including specificity, sensitivity, and AUC, are calculated on the test set (20%) for model validation. Feature importance scores are computed for each baseline variable, providing a ranked analysis of each immune parameter’s contribution to LAIV response prediction. (C) Multiclass ROC plot of the gradient boosting machine model evaluated on the test set (20%), displaying predictive accuracy across all 3 classes: CD8+ T cell responders (green); mucosal responders (orange); and systemic, broad influenza A responders (purple) in a one-versus-all comparison. (D) Variable importance score table for the gradient boosting machine model, showcasing the cumulative importance of the selected baseline features across the 3 predicted classes, highlighting the most influential parameters in LAIV immunogenicity prediction.