❌

Normal view

Received β€” 13 September 2026 ⏭

A Mitochondrial-Related Gene Signature for Diagnosis and Immune Microenvironment Modulation in Lung Cancer and Venous Thromboembolism

βœ‡(Multiomics OR Omics) AND (Lung OR gastric OR Hepatocellular)
By: Xue Li Zhang Β· Ai Li Gao Β· Zhan Ju Liu
12 September 2026 at 18:00

World J Oncol. 2026 Sep 4;17(5):683-704. doi: 10.14740/wjon2815. eCollection 2026 Oct.

ABSTRACT

BACKGROUND: Lung cancer (LC) and venous thromboembolism (VTE) are closely associated, with VTE contributing to morbidity and mortality among patients with LC. We aimed to identify and characterize a mitochondrial-related transcriptomic signature shared between LC and VTE and to explore its association with immune microenvironment features.

METHODS: We applied a multiomics approach focused on mitochondrial-related signaling pathways. Publicly available transcriptomic datasets were analyzed using differential expression profiling and weighted gene co-expression network analysis to identify key regulatory genes. These genes were intersected with a mitochondrial gene set and subjected to functional enrichment analysis. Least absolute shrinkage and selection operator (LASSO) regression was used to identify candidate diagnostic genes validation. Immune cell infiltration was quantified, and associated regulatory mechanisms were explored.

RESULTS: Thirty-nine shared crosstalk genes were identified and were primarily enriched in mitochondrial metabolic processes. LASSO regression identified a five-gene candidate signature (ACAA1, HSD17B10, MTIF2, THOP1, and PDE2A). The model exhibited promising discriminatory performance (area under the curve > 0.9 in LC dataset and 0.7-0.9 in VTE dataset). These genes were significantly dysregulated and were associated with altered immune cell infiltration, particularly in dendritic cell and T cell subsets.

CONCLUSION: We identified a mitochondrial-related gene signature reflecting shared transcriptomic correlates between LC and VTE. The signature showed variable performance across disease contexts and correlative associations with immune features, supporting its role as a candidate biomarker for further investigation. Prospective validation in independent clinical cohorts is required before any translational application.

PMID:42730163 | PMC:PMC13568737 | DOI:10.14740/wjon2815

Received β€” 2 April 2026 ⏭

CoMaTrack: Competitive Multi-Agent Game-Theoretic Tracking with Vision-Language-Action Models

βœ‡cs.AI, q-bio.NC updates on arXiv.org
By: Youzhi Liu Β· Li Gao Β· Liu Liu Β· Mingyang Lv Β· Yang Cai
1 April 2026 at 12:00
arXiv:2603.22846v2 Announce Type: replace Abstract: Embodied Visual Tracking (EVT), a core dynamic task in embodied intelligence, requires an agent to precisely follow a language-specified target. Yet most existing methods rely on single-agent imitation learning, suffering from costly expert data and limited generalization due to static training environments. Inspired by competition-driven capability evolution, we propose CoMaTrack, a competitive game-theoretic multi-agent reinforcement learning framework that trains agents in a dynamic adversarial setting with competitive subtasks, yielding stronger adaptive planning and interference-resilient strategies. We further introduce CoMaTrack-Bench, the first open-source Habitat-based benchmark protocol and episode set for language-conditioned competitive EVT featuring dynamic dueling, featuring game scenarios between a tracker and adaptive opponents across diverse environments and instructions, enabling standardized robustness evaluation under active adversarial interactions. Experiments show that CoMaTrack achieves state-of-the-art results on both standard benchmarks and CoMaTrack-Bench. Notably, a 3B VLM trained with our framework surpasses previous single-agent imitation learning methods based on 7B models on the challenging EVT-Bench, achieving 92.1% in STT, 74.2% in DT, and 57.5% in AT. The benchmark code will be available at https://github.com/wlqcode/CoMaTrack-Bench.
Received β€” 26 March 2026 ⏭

CoMaTrack: Competitive Multi-Agent Game-Theoretic Tracking with Vision-Language-Action Models

βœ‡cs.AI, q-bio.NC updates on arXiv.org
By: Youzhi Liu Β· Li Gao Β· Liu Liu Β· Mingyang Lv Β· Yang Cai
25 March 2026 at 12:00
arXiv:2603.22846v1 Announce Type: new Abstract: Embodied Visual Tracking (EVT), a core dynamic task in embodied intelligence, requires an agent to precisely follow a language-specified target. Yet most existing methods rely on single-agent imitation learning, suffering from costly expert data and limited generalization due to static training environments. Inspired by competition-driven capability evolution, we propose CoMaTrack, a competitive game-theoretic multi-agent reinforcement learning framework that trains agents in a dynamic adversarial setting with competitive subtasks, yielding stronger adaptive planning and interference-resilient strategies. We further introduce CoMaTrack-Bench, the first benchmark for competitive EVT, featuring game scenarios between a tracker and adaptive opponents across diverse environments and instructions, enabling standardized robustness evaluation under active adversarial interactions. Experiments show that CoMaTrack achieves state-of-the-art results on both standard benchmarks and CoMaTrack-Bench. Notably, a 3B VLM trained with our framework surpasses previous single-agent imitation learning methods based on 7B models on the challenging EVT-Bench, achieving 92.1% in STT, 74.2% in DT, and 57.5% in AT. The benchmark code will be available at https://github.com/wlqcode/CoMaTrack-Bench
There are no more articles
❌