基于生物信息学和机器学习的肺结核特征基因筛选 及其免疫细胞浸润特征分析

打开文本图片集

中图分类号：R734.2；R521；TP181

文献标识码：A

DOI：10.3969/j.issn.1006-1959.2026.09.002

文章编号：1006-1959（2026）09-0014-08

Screening of Tuberculosis Characteristic Genes and Analysis of its Immune Cell Infiltration Characteristics Based on Bioinformatics and Machine Learning

HUANG Qi， HE Jiaxin， MA Zhuanzhuan， Zainur·Yomair

（School of Medicine， Xinjiang College of Science & Technology， Korla ， Xinjiang， China）

Abstract： Objective To identify key signature genes of tuberculosis and explore its immune cell infiltration characteristics using bioinformatics and machine learning methods. Methods Data were obtained from the GEO database for differential expression gene （DEG） analysis and GO/KEGG functional enrichment analysis. LASSO regression and SVM-RFE were used to select signature genes， and ROC curves were plotted to evaluate their diagnostic value. Immune infiltration characteristics were analyzed using the CIBERSORT algorithm， and the correlation between signature genes and immune cells was examined. Finally， an miRNA-mRNA regulatory network related to TB was constructed. Results A total of 225 DEGs were identified. GO analysis showed enrichment in biological processes such as immune response activation， monocyte differentiation， and cytokine receptor binding， while KEGG analysis highlighted pathways including the NOD-like receptor signaling pathway and Th17 cell differentiation. Five signature genes were selected through machine learning， among which GBP5 and MUC1 showed significant differential expression in the validation set （P<0.01）， with ROC AUC value >0.8， both two genes were significantly correlated with immune cell infiltration （P<0.05）. Conclusion GBP5 and MUC1 are the key characteristic genes of tuberculosis. The constructed tuberculosis-related miRNA-mRNA regulatory network is helpful to provide theoretical support for the early diagnosis， treatment intervention and vaccine development of tuberculosis.

Key words： Tuberculosis; Key genes; Immune cell infiltration; Machine learning

肺结核是由结核分枝杆菌引起的全球性传染病，至今仍然是导致人类死亡的主要原因之一。（剩余9191字）