ImmuneNet: Composition-Aware Quantification of Adaptive Lymphocytes in High-Grade Serous Ovarian Cancer

Jaelyn S. Liang

Keywords: Digital cytometry, virtual instrument, bulk RNA-seq, adaptive lymphocytes, tumor-matched single-cell reference, marker curation, pseudo-bulk simulation, composition-aware evaluation, isotonic regression, temperature scaling, gene-wise attention, ridge regularization, ovarian cancer, HGSOC, domain shift

TL;DR: ImmuneNet is a context-aware, calibrated deconvolution model that leverages tumor-matched single-cell references and curated markers to accurately estimate T and B cell fractions from bulk RNA-seq in ovarian cancer.

Submission Number: 37