Identification and Experimental Validation of Programmed Cell Death-Related Key Genes in Gout Using Bioinformatics Analysis and Machine Learning Based on GEO Database
10.3969/j.issn.1671-7414.2025.04.017
- VernacularTitle:基于GEO数据库采用生物信息学分析和机器学习筛选痛风中程序性细胞死亡相关关键基因及实验验证
- Author:
Junjie CAO
1
;
Erchuan ZHAO
;
Rui WANG
;
Jiaojiao LIU
Author Information
1. 西安市第五医院检验科,西安 710082
- Publication Type:Journal Article
- Keywords:
gout;
programmed cell death;
immune infiltration;
bioinformatics analysis;
machine learning
- From:
Journal of Modern Laboratory Medicine
2025;40(4):97-104
- CountryChina
- Language:Chinese
-
Abstract:
Objective To screening and validation of key genes for programmed cell death(PCD)in gout through bioinformatics and machine learning and immunoinfiltration analysis.Methods Gout-related datasets were obtained from the Gene Expression Omnibus(GEO)database,comprising the human gout dataset GSE160170 and murine gout dataset GSE190138,which served as the training cohort.Differentiall expression genes(DEGs)were screened with difference factor>2 and P<0.05.The DEGs of two data sets were intersected to obtain the common DEGs(co-DEGs).The co-DEGs were enriched by GO function and KEGG pathway analysis.The combination of co-DEGs and PCD related gene set was used to obtain PCD related DEGs.A PPI network was built in the STRING database.The key module genes were screened in Cytoscape's MCODE,the hub genes were screened in 12 algorithms built into the Cytohubba plugin,including Degree,MCC,DMNC,MCN,EPC,BottleNeck,EcCentricity,closness,Radiality,Betweeness,Stress and Clusteringcoefficoent,the common genes of the two was as candidate genes.The regression model of least absolute shrinkage and selection operator(LASSO)was used to screen key genes in the GSE160170 and GSE190138 data sets respectively,and the intersection of the two was adopted to obtain key genes.The diagnostic value of key genes in gout was evaluated by receiver operating characteristic(ROC)curve.The expression difference of gout related immune cells was investigated by single sample gene set enrichmemt analysis(ssGSEA)immunoinfiltration analysis.Finally,blood samples from 30 gout patients admitted to the Department of Rheumatology,Xi'an Fifth Hospital from February to April 2024 were collected as the experimental group,while blood samples from 30 healthy subjects were collected as the control group.RNA was extracted from the Peripheral blood mononuclear cell(PBMC).Quantitative real time polymerase chain reaction(RT-qPCR)was used to validate the expression of key genes in clinical samples.Results 53 common DEGs of GSE160170 and GSE190138 were obtained,among which 43 genes were up-regulated and 10 were down-regulated.GO and KEGG indicated that most genes were involved in cell death,apoptosis,interlenkin(IL)-17 signaling pathway,tumor necrosis facter(TNF)signaling pathway and nucleotide-binding oligomerization domain-(NOD)-like receptor signaling pathway.12 co-DEGs of programmed cell death and gout were obtained.A total of 7 candidate genes were screened.LASSO regression model screened 5 genes and 4 genes respectively in two datasets,and 3 key genes were abtained by the intersection of the two datasets,which were IL-6,plasminogen activator urokinase receptor(PLAUR)and NOD-like receptor thermal protein domain associated protein3(NLRP3).Validation within the training set revealed that all three genes attained perfect diagnostic performance for gout,with area under the ROC(AUC-ROC)curve values of 1.00.Immunoinfiltration analysis showed that the changes of activated CD4+T cells,activated CD8+T cells and natural killer cells were closely related to the occurrence and development of gout.In the clinical samples,compared with the control group,PLAUR,NLRP3 and IL-6 were highly expressed in gout patients,and the differences were statistically significant(t=18.852,9.633,8.293,all P＜0.05).Conclusion IL-6,PLAUR and NLRP3 provide potential biomarkers and therapeutic targets for the diagnosis and treatment of gout,offering new directions in this field.
