CORTEXA
← Browse
crossrefFrontiers in Pharmacology2026-07-10Cited by 0

Deciphering tuberculosis pathway mechanisms via graph neural networks and multimodal deep learning: a comprehensive AI-driven framework for precision medicine

Yuanli Yang, Huanqing Liu, Qian Lei, Tingting Li

Background Tuberculosis (TB) remains a global health crisis, with complex molecular mechanisms that are not fully understood. Traditional pathway analysis methods fail to capture the intricate non-linear relationships within biological networks. Methods We developed a novel artificial intelligence framework integrating graph neural networks (GNNs), transformer architectures, and multimodal deep learning to decipher TB pathway mechanisms. Our approach constructs a comprehensive pathway-gene interaction network from three critical pathways (Tuberculosis hsa05152, Antigen processing hsa04612, and NF-κB signaling hsa04064) and employs three interconnected models: (1) a Graph Convolutional Network for learning pathway-gene relationships, (2) a Transformer encoder for pathway activity prediction, and (3) a multimodal fusion model with attention mechanisms integrating transcriptomic, pathway, and clinical data. The framework was trained and validated on 467 clinical samples and 529 transcriptomic samples from five GEO datasets. Results The Transformer model achieved strong performance in pathway activity prediction ( R 2 = 0.97, MSE = 0.014), demonstrating high accuracy in capturing pathway activation patterns. The multimodal fusion model achieved strong predictive performance (accuracy 88.2%, AUC-ROC 0.90) in clinical outcome prediction, with attention analysis revealing adaptive weighting of different data modalities. Network analysis identified 27 shared genes between Tuberculosis and Antigen processing pathways, and 18 shared genes between Tuberculosis and NF-κB pathways, indicating coordinated immune regulation. Key pathway-gene interactions were identified, including critical roles of IFNG, TNF, IL1B, and NF-κB signaling components. Conclusion This study represents the first comprehensive application of GNNs and multimodal deep learning to TB pathway analysis. Our framework provides novel insights into TB pathogenesis, identifies potential therapeutic targets, and demonstrates the power of AI-driven approaches for understanding complex disease mechanisms. The interpretability of our models through attention mechanisms enables translation of computational findings into actionable biological insights, with significant implications for precision medicine and personalized TB treatment strategies.

View free PDFSource page

Related papers

openalexFrontiers in Pharmacology2026-07-24

Editorial: Exploring the interplay between stress, neuropeptides, and alcohol use disorder

Kabirullah Lutfy, Marcelo López

Alcohol use disorders (AUD) are major public health issues, with a significant negative impact on the health system and the economy. Unfortunately, only a handful of pharmacotherapeutic agents are available to treat patients with AUD. Notably, these medications have limited effic…

View free PDFSource page
openalexFrontiers in Pharmacology2026-07-24

Editorial: Increasing importance of patients-generated real world data for healthcare policy decisions about medicinal products: volume III

Wai Yee Choon, K C Lee, Paul Scuffham, Shu Chuen Li, David Bin-Chia Wu, Jeff J. Guo

Increasing importance of patient-generated real-world data for healthcare policy decisions about medicinal products: Volume III Real-world data (RWD) and the real-world evidence (RWE) derived from it have moved from the margins to the centre of decision-making about medicinal pro…

View free PDFSource page
openalexFrontiers in Pharmacology2026-07-24

GPR55 and PINK1 as emerging therapeutic targets in Glioma: insights into their modulation by phytochemicals and plant extracts

Andrés David Turizo Smith, Jean Carlos Martinez Perez, Juan Camilo Marín Loaiza, Gonzalo Arboleda

The expression of certain genes and proteins in brain tumors, particularly gliomas, presents potential therapeutic targets and biomarkers. One of such genes encodes the protein PINK1 (PTEN-induced Kinase 1), which has been primarily studied due to its mutations being linked to Pa…

View free PDFSource page
openalexFrontiers in Pharmacology2026-07-24

Indobufen-based dual antiplatelet therapy in patients undergoing PCI: a systematic review and meta-analysis

Yiheng Liu, B Y Liu, Weiran Dai

Background Conventional dual antiplatelet therapy (DAPT) post percutaneous coronary intervention (PCI) comprises aspirin and a P2Y 12 inhibitor. However, aspirin increases gastrointestinal and bleeding risks. Indobufen, a reversible COX-1 inhibitor, may offer a safer alternative.…

View free PDFSource page
openalexFrontiers in Pharmacology2026-07-24

Protective effects and possible mechanisms of matrine-type alkaloids (matrine and oxymatrine) in animal models of ulcerative colitis: a preclinical systematic review and meta-analysis

吕书晴, H W Yang, Shiying Wang, Jiahui Ye, S H. M. Huang, Linkun Cai

Background Matrine and oxymatrine are bioactive alkaloids from Sophora species with reported anti-inflammatory and antioxidant effects in ulcerative colitis (UC). However, their overall efficacy has not been systematically evaluated. Methods Databases including PubMed, EMBASE, We…

View free PDFSource page
openalexFrontiers in Pharmacology2026-07-23

Immunomodulatory properties of venom-derived crotamine

Isadora Oliveira, I M P L V O Ferreira, Lorena Sousa, João Alberghini-dos-Santos, Felipe Cerni, Manuela Pucca

Crotamine is a low-molecular-weight cationic polypeptide isolated from Crotalus durissus terrificus snake venom, classified as a non-enzymatic toxin with distinctive structural and functional properties. This toxin also exhibits characteristics of a cell-penetrating peptide, owin…

View free PDFSource page