Biologically Disentangled Multi-Omic Modeling Reveals Mechanistic Insights into Pan-Cancer Immunotherapy Resistance

TL;DR

This paper introduces BDVAE, a deep generative model that integrates multi-omics data to predict immunotherapy response and uncover resistance mechanisms across multiple cancer types, providing interpretable biological insights.

Contribution

The study presents a novel modular variational autoencoder that effectively models multi-omics data with biological structure, improving prediction accuracy and interpretability in pan-cancer immunotherapy resistance.

Findings

BDVAE predicts treatment response with AUC-ROC of 0.94.

Identifies resistance mechanisms like immune suppression and metabolic shifts.

Reveals resistance as a biological spectrum rather than binary states.

Abstract

Immune checkpoint inhibitors (ICIs) have transformed cancer treatment, yet patient responses remain highly variable, and the biological mechanisms underlying resistance are poorly understood. While machine learning models hold promise for predicting responses to ICIs, most existing methods lack interpretability and do not effectively leverage the biological structure inherent to multi-omics data. Here, we introduce the Biologically Disentangled Variational Autoencoder (BDVAE), a deep generative model that integrates transcriptomic and genomic data through modality- and pathway-specific encoders. Unlike existing rigid, pathway-informed models, BDVAE employs a modular encoder architecture combined with variational inference to learn biologically meaningful latent features associated with immune, genomic, and metabolic processes. Applied to a pan-cancer cohort of 366 patients across four cancer types treated with ICIs, BDVAE accurately predicts treatment response (AUC-ROC = 0.94 on unseen test data) and uncovers critical resistance mechanisms, including immune suppression, metabolic shifts, and neuronal signaling. Importantly, BDVAE reveals that resistance spans a continuous biological spectrum rather than strictly binary states, reflecting gradations of tumor dysfunction. Several latent features correlate with survival outcomes and known clinical subtypes, demonstrating BDVAE's capability to generate interpretable, clinically relevant insights. These findings underscore the value of biologically structured machine learning in elucidating complex resistance patterns and guiding precision immunotherapy strategies.