Machine learning and weighted gene co-expression network analysis identify a three-gene signature to diagnose rheumatoid arthritis.

Rheumatoid arthritis (RA) is a systemic immune-related disease characterized by synovial inflammation and destruction of joint cartilage. The pathogenesis of RA remains unclear, and diagnostic markers with high sensitivity and specificity are needed urgently. This study aims to identify potential biomarkers in the synovium for diagnosing RA and to investigate their association with immune infiltration. We downloaded four datasets containing 51 RA and 36 healthy synovium samples from the Gene Expression Omnibus database. Differentially expressed genes were identified using R. Then, various enrichment analyses were conducted. Subsequently, weighted gene co-expression network analysis (WGCNA), random forest (RF), support vector machine-recursive feature elimination (SVM-RFE), and least absolute shrinkage and selection operator (LASSO) were used to identify the hub genes for RA diagnosis. Receiver operating characteristic curves and nomogram models were used to validate the specificity and sensitivity of hub genes. Additionally, we analyzed the infiltration levels of 28 immune cells in the expression profile and their relationship with the hub genes using single-sample gene set enrichment analysis. Three hub genes, namely, ribonucleotide reductase regulatory subunit M2 (RRM2), DLG-associated protein 5 (DLGAP5), and kinesin family member 11 (KIF11), were identified through WGCNA, LASSO, SVM-RFE, and RF algorithms. These hub genes correlated strongly with T cells, natural killer cells, and macrophage cells as indicated by immune cell infiltration analysis. RRM2, DLGAP5, and KIF11 could serve as potential diagnostic indicators and treatment targets for RA. The infiltration of immune cells offers additional insights into the underlying mechanisms involved in the progression of RA.

Wu YK ，Liu CD ，Liu C ，Wu J ，Xie ZG ... -  《Frontiers in Immunology》

Identification of diagnostic genes and drug prediction in metabolic syndrome-associated rheumatoid arthritis by integrated bioinformatics analysis, machine learning, and molecular docking.

Interactions between the immune and metabolic systems may play a crucial role in the pathogenesis of metabolic syndrome-associated rheumatoid arthritis (MetS-RA). The purpose of this study was to discover candidate biomarkers for the diagnosis of RA patients who also had MetS. Three RA datasets and one MetS dataset were obtained from the Gene Expression Omnibus (GEO) database. Differential expression analysis, weighted gene co-expression network analysis (WGCNA), and machine learning algorithms including Least Absolute Shrinkage and Selection Operator (LASSO) regression and Random Forest (RF) were employed to identify hub genes in MetS-RA. Enrichment analysis was used to explore underlying common pathways between MetS and RA. Receiver operating characteristic curves were applied to assess the diagnostic performance of nomogram constructed based on hub genes. Protein-protein interaction, Connectivity Map (CMap) analyses, and molecular docking were utilized to predict the potential small molecule compounds for MetS-RA treatment. qRT-PCR was used to verify the expression of hub genes in fibroblast-like synoviocytes (FLS) of MetS-RA. The effects of small molecule compounds on the function of RA-FLS were evaluated by wound-healing assays and angiogenesis experiments. The CIBERSORT algorithm was used to explore immune cell infiltration in MetS and RA. MetS-RA key genes were mainly enriched in immune cell-related signaling pathways and immune-related processes. Two hub genes (TYK2 and TRAF2) were selected as candidate biomarkers for developing nomogram with ideal diagnostic performance through machine learning and proved to have a high diagnostic value (area under the curve, TYK2, 0.92; TRAF2, 0.90). qRT-PCR results showed that the expression of TYK2 and TRAF2 in MetS-RA-FLS was significantly higher than that in non-MetS-RA-FLS (nMetS-RA-FLS). The combination of CMap analysis and molecular docking predicted camptothecin (CPT) as a potential drug for MetS-RA treatment. In vitro validation, CPT was observed to suppress the cell migration capacity and angiogenesis capacity of MetS-RA-FLS. Immune cell infiltration results revealed immune dysregulation in MetS and RA. Two hub genes were identified in MetS-RA, a nomogram for the diagnosis of RA and MetS was established based on them, and a potential therapeutic small molecule compound for MetS-RA was predicted, which offered a novel research perspective for future serum-based diagnosis and therapeutic intervention of MetS-RA.

Huang Y ，Yue S ，Qiao J ，Dong Y ，Liu Y ，Zhang M ，Zhang C ，Chen C ，Tang Y ，Zheng J ... -  《Frontiers in Immunology》

Identification of four-gene signature to diagnose osteoarthritis through bioinformatics and machine learning methods.

Although osteoarthritis (OA) is one of the most prevalent joint disorders, effective biomarkers to diagnose OA are still unavailable. This study aimed to acquire some key synovial biomarkers (hub genes) and analyze their correlation with immune infiltration in OA. Gene expression profiles and clinical characteristics of OA and healthy synovial samples were retrieved from the Gene Expression Omnibus (GEO) database. Hub genes for OA were mined based on a combination of weighted gene co-expression network analysis (WGCNA), the least absolute shrinkage and selection operator (LASSO), support vector machine recursive feature elimination (SVM-RFE), and random forest (RF) algorithms. A diagnostic nomogram model for OA prediction was developed based on the hub genes. Receiver operating characteristic curves (ROC) were performed to confirm the abnormal expression of hub genes in the experimemtal and validation datasets. qRT-PCR using patients' samples were conducted as well. In addition, the infiltration level of 28 immune cells in the expression profile and their relationship with hub genes were analyzed using single-sample GSEA (ssGSEA). 4 hub genes (ZBTB16, TNFSF11, SCRG1 and KDELR3) were obtained by WGCNA, lasso, SVM-RFE, RF algorithms as potential biomarkers for OA. The immune infiltration analyses revealed that hub genes were most correlated with regulatory T cell and natural killer cell. A machine learning model to diagnose OA based on ZBTB16, TNFSF11, SCRG1 and KDELR3 using synovial tissue was constructed, providing theoretical foundation and guideline for diagnostic and treatment targets in OA.

Chen Z ，Wang W ，Zhang Y ，Xue X ，Hua Y ... -  《-》

Identifying hub circadian rhythm biomarkers and immune cell infiltration in rheumatoid arthritis.

Rheumatoid arthritis (RA) is a chronic systemic autoimmune disease with symptoms characterized by typical circadian rhythmic changes. This study aimed to identify the hub circadian rhythm genes (CRGs) in RA and explore their association with immune cell infiltration and pathogenesis of RA. The differentially expressed CRGs (DECRGs) between RA and normal control samples were screened from Datasets GSE12021 and GSE55235. Gene Ontology, Kyoto Encyclopedia of Genes and Genomes, and Gene Set Enrichment Analysis were used to explore the potential functional mechanisms of DECRGs in RA. Weighted Gene Co-expression Network Analysis and Least Absolute Shrinkage and Selection Operator regression analysis were performed to identify hub CRGs of RA. CIBERSORT was conducted to compare the infiltration level of immune cells in RA and control synovial tissue and their relationship with hub genes. In addition, the diagnostic value of hub biomarkers was evaluated by the area under the receiver operator characteristic curve. Further, a nomogram prediction model was constructed and its significance for clinical decision-making was evaluated. The green module was identified as the hub module associated with RA. Four hub CRGs (EGR1, FOSL2, GADD45B, and NFIL3) were identified and showed that they had the highest specificity and sensitivity for RA diagnosis, respectively. The expression levels and diagnostic values of these genes were externally validated in the dataset GSE55457. A nomogram prediction model based on the four hub CRGs was constructed and proved to have a certain clinical decision value. Additionally, the correlation analysis of immune cells with hub genes showed that all hub genes were significantly positively correlated with activated mast cells, resting memory CD4+ T cells, and monocytes. Whereas, all hub genes were negatively correlated with plasma cells, CD8+ T cells, and activated memory CD4+ T cells. Meanwhile, FOSL2 and GADD45B were negatively correlated with Tfh cells. Four hub CRGs were identified and showed excellent diagnostic value for RA. These genes may be involved in the pathological process of RA by disrupting the rhythmic oscillations of cytokines through immune-related pathways and could be considered molecular targets for future chronotherapy against RA.

Wen P ，Ma T ，Zhang B ，Hao L ，Wang Y ，Guo J ，Song W ，Wang J ，Zhang Y ... -  《Frontiers in Immunology》

Machine learning to identify immune-related biomarkers of rheumatoid arthritis based on WGCNA network.

This study was designed to identify the potential diagnostic biomarkers of rheumatoid arthritis (RA) and to explore the potential pathological relevance of immune cell infiltration in this disease. Three previously published datasets containing gene expression data from 35 RA patients and 29 controls (GSE55235, GSE55457, and GSE12021) were downloaded from the GEO database, after which a weighted correlation network analysis (WGCNA) approach was utilized to clarify differentially abundant genes. Candidate biomarkers of RA were then identified via the use of a LASSO regression model and support vector machine recursive feature elimination (SVM-RFE) analyses. Data were validated based upon the area under the receiver operating characteristic curve (AUC) values, with hub genes being identified as those with an AUC > 85% and a P value < 0.05. Lastly, the CIBERSORT algorithm was used to assess immune cell infiltration of RA tissues, and correlations between immune cell infiltration and disease-related diagnostic biomarkers were assessed. The green-yellow module containing 87 genes was found to be highly correlated with RA positivity. FADD, CXCL2, and CXCL8 were identified as potential RA diagnostic biomarkers (AUC > 0.85), and these results were validated using the GSE77298 dataset. Immune cell infiltration analyses revealed the expression of hub genes to be correlated with mast cells, monocytes, activated NK cells, CD8 T cells, resting dendritic cells, and plasma cells. These data indicate that FADD, CXCL2, and CXCL8 are valuable diagnostic biomarkers of RA, offering new insight that can guide future studies of RA incidence and progression.

Chen Y ，Liao R ，Yao Y ，Wang Q ，Fu L ... -  《-》