AUTHOR=Pan Bao-ze , Jiang Ming-jun , Deng Li-ming , Chen Jie , Dai Xian-peng , Wu Zi-xuan , Deng Zhi-he , Luo Dong-yang , Wang Yang-yi-jing , Ning Dan , Xiong Guo-zuo , Bi Guo-shan 

TITLE=Integration of bulk RNA-seq and scRNA-seq reveals transcriptomic signatures associated with deep vein thrombosis

JOURNAL=Frontiers in Genetics

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2025.1551879

DOI=10.3389/fgene.2025.1551879

ISSN=1664-8021

ABSTRACT=BackgroundDeep vein thrombosis (DVT) is a prevalent peripheral vascular disease. The intricate and multifaceted nature of the associated mechanisms hinders a comprehensive understanding of disease-relevant targets. This study aimed to identify and examine the most distinctive genes linked to DVT.MethodsIn this study, the bulk RNA sequencing (bulk RNA-seq) analysis was conducted on whole blood samples from 11 DVT patients and six control groups. Topology analysis was performed using seven protein-protein interaction (PPI) network algorithms. The combination of weighted correlation network analysis (WGCNA) and clinical prediction models was employed to validate hub DEGs. Furthermore, single-cell RNA sequencing (scRNA-seq) was performed on peripheral blood samples from 3 DVT patients and three control groups to probe the cellular localization of target genes. Based on the same methodology as the internal test set, 12 DVT patients and six control groups were collected to construct an external test set and validated using machine learning (ML) algorithms and immunofluorescence (IF). Concurrently, the examination of the pathways in disparate cell populations was conducted on the basis of the CellChat pathway.ResultsA total of 193 DEGs were identified in the internal test set. Additionally, a total of eight highly characteristic genes (including TLR1, TLR7, TLR8, CXCR4, DDX58, TNFSF10, FCGR1A and CD36) were identified by the PPI network algorithm. In accordance with the WGCNA model, the aforementioned genes were all situated within the blue core module, exhibiting a correlation coefficient of 0.84. The model demonstrated notable disparities in TLR8 (P = 0.018, AUC = 0.847), CXCR4 (P = 0.00088, AUC = 1.000), TNFSF10 (P = 0.00075, AUC = 0.958), and FCGR1A (P = 0.00022, AUC = 0.986). Furthermore, scRNA-seq demonstrated that B cells, T cells and monocytes play an active role in DVT. In the external validation set, CXCR4 was validated as a potential target by the ML algorithm and IF. In the context of the CellChat pathway, it indicated that MIF - (CD74 + CXCR4) plays a potential role.ConclusionThe findings of this study indicate that CXCR4 may serve as a potential genetic marker for DVT, with MIF - (CD74 + CXCR4) potentially implicated in the regulatory mechanisms underlying DVT.