Al-Humadi, R. | Search bioRxiv

bioRxiv2026

Deep-learning-driven tumor microenvironment profiling improves immunotherapy response prediction

Immunotherapy has transformed the management of many cancers, such as clear cell renal cell carcinoma (ccRCC), yet durable clinical benefit is achieved in only a subset of patients. Immunotherapy outcomes are governed by the multicellular tumor microenvironment (TME), through the structured and directional interplay of regulatory programs and ligand-receptor communications. Yet most existing methods for immunotherapy response prediction rely on independent signatures or cell types and fail to model the structured and directional signaling network within the TME. To address this gap, we introduce EcoNet, a computational framework that first models ecosystems consisting of co-occurring cell states, then constructs intercellular and intracellular signaling networks based on the ecosystems, and applies a graph attention network to derive embeddings from the signaling network for immunotherapy response prediction. Pretrained on 2,532 pan-cancer immunotherapy samples across multiple cohorts, EcoNet outperformed existing prediction methods by 7.5-19.9% and marker-based signatures by 4.4-21.2% in AUC, and its high-attention genes were enriched among those required for T-cell-mediated tumor killing in an independent CRISPR screen; applied to ccRCC with disease-specific fine-tuning, it stratified an independent ccRCC cohort by clinical outcome (P=0.024). Biologically, high-attention genes capture an interferon-responsive program that enhances antigen presentation and immune infiltration. To validate these findings, we curated a ccRCC clinical cohort profiled with CosMx single-cell spatial transcriptomics (46 samples, ~450,000 cells) and found the response-associated TME configuration identified by EcoNet localized to immune-enriched niches, supporting the response stratification. These results establish EcoNet as a generalizable framework for TME-informed therapeutic response prediction, readily extensible to other cancer types and therapeutic contexts.

bioinformatics↗

bioRxiv2026

Mouse lymph node colonization models predict human tumor metastasis in melanoma

Aggressive tumor progression and distant metastases of cancers such as cutaneous melanoma are thought to be preceded by regional lymph node (LN) infiltration. Previous mouse studies have selectively enriched for tumor cells with enhanced LN migratory capacity but the mechanistic biology underpinning this proclivity has not been directly linked to human outcomes. Likewise, the exact role LN colonization plays in distant metastasis in humans remains unclear. To translate and conserve findings from mouse models to humans, we used a computational framework termed Translatable Components Regression (TransComp-R) to integrate mouse LN tumor samples with human primary and metastatic tumors. We identified transcriptional programs of early- and late-stage mouse tumor generations that stratified primary and metastatic tumors from patients with melanoma. These transcriptional programs associated with immune function, cell cycle checkpoints, and DNA-to-protein biosynthesis pathways. We then employed EcoTyper and identified six distinct melanoma-specific cellular communities (ecotypes) comprising co-occurring cell states. Of these six melanoma ecotypes, we identified a community strongly associated with primary tumor cells, while the other five represented a spectrum of metastatic tumor microenvironments with increased adverse outcomes. These ecotypes and their significantly expressed genes may contribute to the tumor-immune microenvironment that facilitates LN colonization in melanoma. We further validated the physical co-localization of these melanoma ecotypes using spatial transcriptomics. Our integration of pre-clinical models with human data reveals potential biological pathways and transcriptomic signatures for future investigation and supports the utility of mouse models in studying cancer metastasis.

systems biology↗