Screening of Key Genes for Oosteoporosis in Postmenopausal Osteoporosis with Yin Deficiency Based on Bioinformatics and Machine Learning
Objective To apply bioinformatics and machine learning to screen for key genes in postmenopausal osteoporosis(PMOP)with Yin deficiency.Methods The peripheral blood mononuclear cell data of GSE87474 people with moderate constitution and yin deficiency constitution were downloaded from GEO database,and the data sets GSE56116 and GSE100609 related to PMOP were downloaded and combined with SVA package.Secondly,weighted gene co-expression network was used to screen genes closely related to Yin deficiency in traditional Chinese medicine(TCM).limma package was used to screen the differentially expressed genes in the combined data set.They were intersected to obtain genes associated with Yin deficiency PMOP.Then,the minimum absolute convergence and selection operator(LASSO),support vector machine-recursive feature elimination(SVM-RFE)and random forest(RF)algorithms were used to identify potential key genes in the process of PMOP in Yin deficiency and evaluate their diagnostic effectiveness.Finally,single gene set enrichment analysis(GSEA)was performed for key genes.Results A total of 46 genes related to PMOP with Yin-deficiency constitution were obtained,and one key gene,solute carrier family 39 member 8(SLC39A8),was screened out by using LASSO,SVM-RFE and RF algorithms.Single-gene GSEA showed that the SLC39A8 high expression group was enriched in 104 signaling pathways including HIPPO,interleukin-12,P38MAPK,P53,and apoptosis.Conclusion This study found that SLC39A8 may be the key gene of PMOP in Yin deficiency,and the results can provide a new idea and breakthrough point for further elucidation of the molecular mechanism of PMOP in Yin deficiency and the treatment of PMOP from the constitution theory of TCM.
Postmenopausal osteoporosisConstitution with Yin-deficiencyBioinformaticsMachine learningBiomarkers
万方数据
维普
