IDEAL-Age: an interpretable deep learning framework for single-cell resolution profiling of immunological aging

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Not reported in abstract — specific performance metrics (e.g., mean absolute error, Pearson correlation, R²) are not provided

• Primary Outcome: IDEAL-Age demonstrated superior predictive performance for immunological age at single-cell resolution compared to 35 benchmarked methods across independent cohorts

• Nnt Or Sensitivity: Not applicable (computational/diagnostic framework) — sensitivity, specificity, and AUC for disease state classification are not reported in the abstract

• Confidence Interval: Not reported

Clinical Application

Low feasibility for immediate clinical translation. Single-cell RNA sequencing of PBMCs requires specialised laboratory infrastructure, significant cost (estimated AUD $500–2,000+ per sample for library preparation and sequencing), bioinformatics expertise, and standardised protocols. The IDEAL-Age framework itself requires computational resources and expertise beyond standard clinical laboratory settings. No clinical decision support tool or validated clinical assay has been described. IDEAL-Age has no current direct relevance to PBS-listed therapies or TGA-approved diagnostics. In the Australian context, immunosenescence research is relevant to: (1) the National Immunisation Program, particularly influenza and COVID-19 vaccine responsiveness in older adults; (2) RACGP guidelines for healthy aging and preventive care; (3) rheumatology practice given the SLE application. However, scRNA-seq-based immunological aging profiling is not part of any current Australian clinical pathway. The Australian Genomics initiative and MRFF-funded precision medicine programs may eventually provide infrastructure for such approaches. Ethnic diversity considerations are important given Australia's multicultural population — validation in Indigenous Australian and Asian-Australian cohorts would be necessary before any clinical application. Potentially applicable to research populations undergoing PBMC-based single-cell transcriptomic profiling, particularly in the context of autoimmune disease (SLE), vaccine responsiveness research, and healthy aging studies. Not currently applicable to routine clinical populations.

Abstract

Immunosenescence increases susceptibility to infection and reduces vaccine responsiveness, yet bulk transcriptomic clocks obscure the cellular heterogeneity underlying this process. Here, we present IDEAL-Age, an interpretable deep learning framework that operates directly on single-cell PBMC transcriptomes. Benchmarking against 35 methods across independent cohorts demonstrates superior predictive performance. The framework's interpretability uncovers linear and non-linear gene contribution trajectories that reveal phase-specific physiological transitions, and identifies youth-associated or aging-associated cellular roles. Application to systemic lupus erythematosus reveals accelerated immunological aging driven by interferon-associated monocyte shifts. IDEAL-Age establishes a high-resolution computational framework for deciphering systemic immune aging.

References

1. 1.Xu, Y., Luo, Z., He, K., Zhang, F., Zhang, Y., Wang, J., Wen, H., Li, Y., & Han, D. (2026). IDEAL-Age: an interpretable deep learning framework for single-cell resolution profiling of immunological aging. Genome Biology. https://doi.org/10.1038/s41467-026-70645-3