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

From molecular subgroups to molecular targeted therapy in rheumatoid arthritis: A bioinformatics approach

1Background

Rheumatoid Arthritis (RA) is a heterogeneous autoimmune disease with multiple unidentified pathogenic factors. The inconsistency between molecular subgroups poses challenges for early diagnosis and personalized treatment strategies. In this study, we aimed to accurately distinguish RA patients at the transcriptome level using bioinformatics methods.

2Methods

We collected a total of 362 transcriptome datasets from RA patients in three independent samples from the GEO database. Consensus clustering was performed to identify molecular subgroups, and clinical features were assessed. Differential analysis was employed to annotate the biological functions of specifically upregulated genes between subgroups.

3Results

Based on consensus clustering of RA samples, we identified three robust molecular subgroups, with Subgroup III representing the high-risk subgroup and Subgroup II exhibiting a milder phenotype, possibly associated with relatively higher levels of autophagic ability. Subgroup I showed biological functions mainly related to viral infections, cellular metabolism, protein synthesis, and inflammatory responses. Subgroup II involved autophagy of mitochondria and organelles, protein localization, and organelle disassembly pathways, suggesting heterogeneity in the autophagy process of mitochondria that may play a protective role in inflammatory diseases. Subgroup III represented a high-risk subgroup with pathological processes including abnormal amyloid precursor protein activation, promotion of inflammatory response, and cell proliferation.

4Conclusion

The classification of the RA dataset revealed pathological heterogeneity among different subgroups, providing new insights and a basis for understanding the molecular mechanisms of RA, identifying potential therapeutic targets, and developing personalized treatment approaches.

Identification of steroid-induced osteonecrosis of the femoral head biomarkers based on immunization and animal experiments

BACKGROUND

Steroid-induced osteonecrosis of femoral head (SONFH) is a severe health risk, and this study aims to identify immune-related biomarkers and pathways associated with the disease through bioinformatics analysis and animal experiments.

METHOD

Using SONFH-related datasets obtained from the GEO database, we performed differential expression analysis and weighted gene co-expression network analysis (WGCNA) to extract SONFH-related genes. A protein-protein interaction (PPI) network was then constructed, and core sub-network genes were identified. Immune cell infiltration and clustering analysis of SONFH samples were performed to assess differences in immune cell populations. WGCNA analysis was used to identify module genes associated with immune cells, and hub genes were identified using machine learning. Internal and external validation along with animal experiments were conducted to confirm the differential expression of hub genes and infiltration of immune cells in SONFH.

RESULTS

Differential expression analysis revealed 502 DEGs. WGCNA analysis identified a blue module closely related to SONFH, containing 1928 module genes. Intersection analysis between DEGs and blue module genes resulted in 453 intersecting genes. The PPI network and MCODE module identified 15 key targets enriched in various signaling pathways. Analysis of immune cell infiltration showed statistically significant differences in CD8 + t cells, monocytes, macrophages M2 and neutrophils between SONFH and control samples. Unsupervised clustering classified SONFH samples into two clusters (C1 and C2), which also exhibited significant differences in immune cell infiltration. The hub genes (ICAM1, NR3C1, and IKBKB) were further identified using WGCNA and machine learning analysis. Based on these hub genes, a clinical prediction model was constructed and validated internally and externally. Animal experiments confirmed the upregulation of hub genes in SONFH, with an associated increase in immune cell infiltration.

CONCLUSION

This study identified ICAM1, NR3C1, and IKBKB as potential immune-related biomarkers involved in immune cell infiltration of CD8 + t cells, monocytes, macrophages M2, neutrophils and other immune cells in the pathogenesis of SONFH. These biomarkers act through modulation of the chemokine signaling pathway, Toll-like receptor signaling pathway, and other pathways. These findings provide valuable insights into the disease mechanism of SONFH and may aid in future drug development efforts.

Prediction of immune molecules activity during burn wound healing among elderly patients: in-silico analyses: experimental research

Introduction

Burn injuries lead to dysregulation of immune molecules, impacting cellular and humoral immune pathways. This study aims to determine the prediction of immune molecule activity during burn wound healing among elderly patients.

Methods

The current study utilized the Gene Expression Omnibus (GEO) database to extract the proper gene set. Also, the literature review was conducted in the present study to find immune signatures. The study used the "enrich r" website to identify the biological functions of extracted genes. The critical gene modules related to mortality were identified using the weighted gene co-expression network analysis (WGCNA) R package.

Results

The appreciated GSE was extracted. According to the data, the most upregulated signatures were related to natural killer (NK) cells, and the most downregulated signatures were associated with M1 macrophages. Also, the results of WGCNA have shown that the most related gene modules (P<107 and score 0.17) to mortality were investigated, and the modules 100 first genes were extracted. Additionally, the enrich r analysis has demonstrated related pathways, including the immune process, including regulation of histamine secreted from mast cell (P<0.05), T helper 17 cell differentiation (P<0.05), and autophagy (P<0.05) were obtained. Finally, by network analysis, the critical gene "B3GNT5" were obtained (degree>ten and "betweenness and centrality">30 were considered).

Conclusion

The study identified significant changes in macrophage and NK cell expression patterns post-burn injury, linking them to potential improvements in clinical outcomes and wound healing. The gene B3GNT5, associated with mortality, was highlighted as a key marker for prognostic evaluation.

Unveiling the link between lactate metabolism and rheumatoid arthritis through integration of bioinformatics and machine learning

Rheumatoid arthritis (RA) is a persistent autoimmune condition characterized by synovitis and joint damage. Recent findings suggest a potential link to abnormal lactate metabolism. This study aims to identify lactate metabolism-related genes (LMRGs) in RA and investigate their correlation with the molecular mechanisms of RA immunity. Data on the gene expression profiles of RA synovial tissue samples were acquired from the gene expression omnibus (GEO) database. The RA database was acquired by obtaining the common LMRDEGs, and selecting the gene collection through an SVM model. Conducting the functional enrichment analysis, followed by immuno-infiltration analysis and protein-protein interaction networks. The results revealed that as possible markers associated with lactate metabolism in RA, KCNN4 and SLC25A4 may be involved in regulating macrophage function in the immune response to RA, whereas GATA2 is involved in the immune mechanism of DC cells. In conclusion, this study utilized bioinformatics analysis and machine learning to identify biomarkers associated with lactate metabolism in RA and examined their relationship with immune cell infiltration. These findings offer novel perspectives on potential diagnostic and therapeutic targets for RA.

Machine learning-based identification of novel hub genes associated with oxidative stress in lupus nephritis: implications for diagnosis and therapeutic targets

BACKGROUND

Lupus nephritis (LN) is a complication of SLE characterised by immune dysfunction and oxidative stress (OS). Limited options exist for LN. We aimed to identify LN-related OS, highlighting the need for non-invasive diagnostic and therapeutic approaches.

METHODS

LN-differentially expressed genes (DEGs) were extracted from Gene Expression Omnibus datasets (GSE32591, GSE112943 and GSE104948) and Molecular Signatures Database for OS-associated DEGs (OSEGs). Functional enrichment analysis was performed for OSEGs related to LN. Weighted gene co-expression network analysis identified hub genes related to OS-LN. These hub OSEGs were refined as biomarker candidates via least absolute shrinkage and selection operator. The predictive value was validated using receiver operating characteristic (ROC) curves and nomogram for LN prognosis. We evaluated LN immune cell infiltration using single-sample gene set enrichment analysis and CIBERSORT. Additionally, gene set enrichment analysis explored the functional enrichment of hub OSEGs in LN.

RESULTS

The study identified four hub genes, namely STAT1, PRODH, TXN2 and SETX, associated with OS related to LN. These genes were validated for their diagnostic potential, and their involvement in LN pathogenesis was elucidated through ROC and nomogram. Additionally, alterations in immune cell composition in LN correlated with hub OSEG expression were observed. Immunohistochemical analysis reveals that the hub gene is most correlated with activated B cells and CD8 T cells. Finally, we uncovered that the enriched pathways of OSEGs were mainly involved in the PI3K-Akt pathway and the Janus kinase-signal transducer and activator of transcription pathway.

CONCLUSION

These findings contribute to advancing our understanding of the complex interplay between OS, immune dysregulation and molecular pathways in LN, laying a foundation for the identification of potential diagnostic biomarkers and therapeutic targets.

Identification of important modules and biomarkers in tuberculosis based on WGCNA

Background

Tuberculosis (TB) is a significant public health concern, particularly in China. Long noncoding RNAs (lncRNAs) can provide abundant pathological information regarding etiology and could include candidate biomarkers for diagnosis of TB. However, data regarding lncRNA expression profiles and specific lncRNAs associated with TB are limited.

Methods

We performed ceRNA-microarray analysis to determine the expression profile of lncRNAs in peripheral blood mononuclear cells (PBMCs). Weighted gene co-expression network analysis (WGCNA) was then conducted to identify the critical module and genes associated with TB. Other bioinformatics analyses, including Kyoto Encyclopedia of Genes and Genomes (KEGG), Gene Ontology (GO), and co-expression networks, were conducted to explore the function of the critical module. Finally, real-time quantitative polymerase chain reaction (qPCR) was used to validate the candidate biomarkers, and receiver operating characteristic analysis was used to assess the diagnostic performance of the candidate biomarkers.

Results

Based on 8 TB patients and 9 healthy controls (HCs), a total of 1,372 differentially expressed lncRNAs were identified, including 738 upregulated lncRNAs and 634 downregulated lncRNAs. Among all lncRNAs and mRNAs in the microarray, the top 25% lncRNAs (3729) and top 25% mRNAs (2824), which exhibited higher median expression values, were incorporated into the WGCNA. The analysis generated 16 co-expression modules, among which the blue module was highly correlated with TB. GO and KEGG analyses showed that the blue module was significantly enriched in infection and immunity. Subsequently, considering module membership values (>0.85), gene significance values (>0.90) and fold-change value (>2 or < 0.5) as selection criteria, the top 10 upregulated lncRNAs and top 10 downregulated lncRNAs in the blue module were considered as potential biomarkers. The candidates were then validated in an independent validation sample set (31 TB patients and 32 HCs). The expression levels of 8 candidates differed significantly between TB patients and HCs. The lncRNAs ABHD17B (area under the curve [AUC] = 1.000) and ENST00000607464.1 (AUC = 1.000) were the best lncRNAs in distinguishing TB patients from HCs.

Conclusion

This study characterized the lncRNA profiles of TB patients and identified a significant module associated with TB as well as novel potential biomarkers for TB diagnosis.

Machine learning application in autoimmune diseases: State of art and future prospectives

Autoimmune diseases are a group of disorders resulting from an alteration of immune tolerance, characterized by the formation of autoantibodies and the consequent development of heterogeneous clinical manifestations. Diagnosing autoimmune diseases is often complicated, and the available prognostic tools are limited. Machine learning allows us to analyze large amounts of data and carry out complex calculations quickly and with minimal effort. In this work, we examine the literature focusing on the use of machine learning in the field of the main systemic (systemic lupus erythematosus and rheumatoid arthritis) and organ-specific autoimmune diseases (type 1 diabetes mellitus, autoimmune thyroid, gastrointestinal, and skin diseases). From our analysis, interesting applications of machine learning emerged for developing algorithms useful in the early diagnosis of disease or prognostic models (risk of complications, therapeutic response). Subsequent studies and the creation of increasingly rich databases to be supplied to the algorithms will eventually guide the clinician in the diagnosis, allowing intervention when the pathology is still in an early stage and immediately directing towards a correct therapeutic approach.

Comprehensive analysis revealed the immunoinflammatory targets of rheumatoid arthritis based on intestinal flora, miRNA, transcription factors, and RNA-binding proteins databases, GSEA and GSVA pathway observations, and immunoinfiltration typing

OBJECTIVE

Rheumatoid arthritis (RA) is a chronic inflammatory arthritis. This study aimed to identify potential biomarkers and possible pathogenesis of RA using various bioinformatics analysis tools.

METHODS

The GMrepo database provided a visual representation of the analysis of intestinal flora. We selected the GSE55235 and GSE55457 datasets from the Gene Expression Omnibus database to identify differentially expressed genes (DEGs) separately. With the intersection of these DEGs with the target genes associated with RA found in the GeneCards database, we obtained the DEGs targeted by RA (DERATGs). Subsequently, Disease Ontology, Gene Ontology, and the Kyoto Encyclopedia of Genes and Genomes were used to analyze DERATGs functionally. Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA) were performed on the data from the gene expression matrix. Additionally, the protein-protein interaction network, transcription factor (TF)-targets, target-drug, microRNA (miRNA)-mRNA networks, and RNA-binding proteins (RBPs)-DERATGs correlation analyses were built. The CIBERSORT was used to evaluate the inflammatory immune state. The single-sample GSEA (ssGSEA) algorithm and differential analysis of DERATGs were used among the infiltration degree subtypes.

RESULTS

There were some correlations between the abundance of gut flora and the prevalence of RA. A total of 54 DERATGs were identified, mainly related to immune and inflammatory responses and immunodeficiency diseases. Through GSEA and GSVA analysis, we found pathway alterations related to metabolic regulations, autoimmune diseases, and immunodeficiency-related disorders. We obtained 20 hub genes and 2 subnetworks. Additionally, we found that 39 TFs, 174 drugs, 2310 miRNAs, and several RBPs were related to DERATGs. Mast, plasma, and naive B cells differed during immune infiltration. We discovered DERATGs' differences among subtypes using the ssGSEA algorithm and subtype grouping.

CONCLUSIONS

The findings of this study could help with RA diagnosis, prognosis, and targeted molecular treatment.

Gut microbiota landscape and potential biomarker identification in female patients with systemic lupus erythematosus using machine learning

Objectives

Systemic Lupus Erythematosus (SLE) is a complex autoimmune disease that disproportionately affects women. Early diagnosis and prevention are crucial for women's health, and the gut microbiota has been found to be strongly associated with SLE. This study aimed to identify potential biomarkers for SLE by characterizing the gut microbiota landscape using feature selection and exploring the use of machine learning (ML) algorithms with significantly dysregulated microbiotas (SDMs) for early identification of SLE patients. Additionally, we used the SHapley Additive exPlanations (SHAP) interpretability framework to visualize the impact of SDMs on the risk of developing SLE in females.

Methods

Stool samples were collected from 54 SLE patients and 55 Negative Controls (NC) for microbiota analysis using 16S rRNA sequencing. Feature selection was performed using Elastic Net and Boruta on species-level taxonomy. Subsequently, four ML algorithms, namely logistic regression (LR), Adaptive Boosting (AdaBoost), Random Forest (RF), and eXtreme gradient boosting (XGBoost), were used to achieve early identification of SLE with SDMs. Finally, the best-performing algorithm was combined with SHAP to explore how SDMs affect the risk of developing SLE in females.

Results

Both alpha and beta diversity were found to be different in SLE group. Following feature selection, 68 and 21 microbiota were retained in Elastic Net and Boruta, respectively, with 16 microbiota overlapping between the two, i.e., SDMs for SLE. The four ML algorithms with SDMs could effectively identify SLE patients, with XGBoost performing the best, achieving Accuracy, Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value, and AUC values of 0.844, 0.750, 0.938, 0.923, 0.790, and 0.930, respectively. The SHAP interpretability framework showed a complex non-linear relationship between the relative abundance of SDMs and the risk of SLE, with Escherichia_fergusonii having the largest SHAP value.

Conclusions

This study revealed dysbiosis in the gut microbiota of female SLE patients. ML classifiers combined with SDMs can facilitate early identification of female patients with SLE, particularly XGBoost. The SHAP interpretability framework provides insight into the impact of SDMs on the risk of SLE and may inform future scientific treatment for SLE.

Identification of hub genes and their correlation with immune infiltrating cells in membranous nephropathy: an integrated bioinformatics analysis

BACKGROUND

Membranous nephropathy (MN) is a chronic glomerular disease that leads to nephrotic syndrome in adults. The aim of this study was to identify novel biomarkers and immune-related mechanisms in the progression of MN through an integrated bioinformatics approach.

METHODS

The microarray data were downloaded from the Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) between MN and normal samples were identified and analyzed by the Gene Ontology analysis, the Kyoto Encyclopedia of Genes and Genomes analysis and the Gene Set Enrichment Analysis (GSEA) enrichment. Hub The hub genes were screened and identified by the weighted gene co-expression network analysis (WGCNA) and the least absolute shrinkage and selection operator (LASSO) algorithm. The receiver operating characteristic (ROC) curves evaluated the diagnostic value of hub genes. The single-sample GSEA analyzed the infiltration degree of several immune cells and their correlation with the hub genes.

RESULTS

We identified a total of 574 DEGs. The enrichment analysis showed that metabolic and immune-related functions and pathways were significantly enriched. Four co-expression modules were obtained using WGCNA. The candidate signature genes were intersected with DEGs and then subjected to the LASSO analysis, obtaining a total of 6 hub genes. The ROC curves indicated that the hub genes were associated with a high diagnostic value. The CD4+ T cells, CD8+ T cells and B cells significantly infiltrated in MN samples and correlated with the hub genes.

CONCLUSIONS

We identified six hub genes (ZYX, CD151, N4BP2L2-IT2, TAPBP, FRAS1 and SCARNA9) as novel biomarkers for MN, providing potential targets for the diagnosis and treatment.

Identifying potential biomarkers for non-obstructive azoospermia using WGCNA and machine learning algorithms

Objective

The cause and mechanism of non-obstructive azoospermia (NOA) is complicated; therefore, an effective therapy strategy is yet to be developed. This study aimed to analyse the pathogenesis of NOA at the molecular biological level and to identify the core regulatory genes, which could be utilised as potential biomarkers.

Methods

Three NOA microarray datasets (GSE45885, GSE108886, and GSE145467) were collected from the GEO database and merged into training sets; a further dataset (GSE45887) was then defined as the validation set. Differential gene analysis, consensus cluster analysis, and WGCNA were used to identify preliminary signature genes; then, enrichment analysis was applied to these previously screened signature genes. Next, 4 machine learning algorithms (RF, SVM, GLM, and XGB) were used to detect potential biomarkers that are most closely associated with NOA. Finally, a diagnostic model was constructed from these potential biomarkers and visualised as a nomogram. The differential expression and predictive reliability of the biomarkers were confirmed using the validation set. Furthermore, the competing endogenous RNA network was constructed to identify the regulatory mechanisms of potential biomarkers; further, the CIBERSORT algorithm was used to calculate immune infiltration status among the samples.

Results

A total of 215 differentially expressed genes (DEGs) were identified between NOA and control groups (27 upregulated and 188 downregulated genes). The WGCNA results identified 1123 genes in the MEblue module as target genes that are highly correlated with NOA positivity. The NOA samples were divided into 2 clusters using consensus clustering; further, 1027 genes in the MEblue module, which were screened by WGCNA, were considered to be target genes that are highly correlated with NOA classification. The 129 overlapping genes were then established as signature genes. The XGB algorithm that had the maximum AUC value (AUC=0.946) and the minimum residual value was used to further screen the signature genes. IL20RB, C9orf117, HILS1, PAOX, and DZIP1 were identified as potential NOA biomarkers. This 5 biomarker model had the highest AUC value, of up to 0.982, compared to other single biomarker models; additionally, the results of this biomarker model were verified in the validation set.

Conclusions

As IL20RB, C9orf117, HILS1, PAOX, and DZIP1 have been determined to possess the strongest association with NOA, these five genes could be used as potential therapeutic targets for NOA patients. Furthermore, the model constructed using these five genes, which possessed the highest diagnostic accuracy, may be an effective biomarker model that warrants further experimental validation.

Systemic lupus erythematosus with high disease activity identification based on machine learning

OBJECTIVE

Clinical evaluation of systemic lupus erythematosus (SLE) disease activity is limited and inconsistent, and high disease activity significantly, seriously impacts on SLE patients. This study aims to generate a machine learning model to identify SLE patients with high disease activity.

METHOD

A total of 1014 SLE patients with low disease activity and 453 SLE patients with high disease activity were included. A total of 94 clinical, laboratory data and 17 meteorological indicators were collected. After data preprocessing, we use mutual information and multisurf to evaluate and select the importance of features. The selected features are used for machine learning modeling. Performance of the model is evaluated and verified by a series of binary classification indicators.

RESULTS

We screened out hematuria, proteinuria, pyuria, low complement, precipitation, sunlight and other features for model construction by integrated feature selection. After hyperparameter optimization, the LGB has the best performance (ROC: AUC = 0.930; PRC: AUC = 0.911, APS = 0.913; balance accuracy: 0.856), and the worst is the naive bayes (ROC: AUC = 0.849; PRC: AUC = 0.719, APS = 0.714; balance accuracy: 0.705). Finally, the selection of features has good consistency in the composite feature importance bar plot.

CONCLUSION

We identify SLE patients with high disease activity by a simple machine learning pipeline, especially the LGB model based on the characteristics of proteinuria, hematuria, pyuria and other feathers screened out by collective feature selection.

Platelets-related signature based diagnostic model in rheumatoid arthritis using WGCNA and machine learning

Background and aim

Rheumatoid arthritis (RA) is an autoinflammatory disease that may lead to severe disability. The diagnosis of RA is limited due to the need for biomarkers with both reliability and efficiency. Platelets are deeply involved in the pathogenesis of RA. Our study aims to identify the underlying mechanism and screening for related biomarkers.

Methods

We obtained two microarray datasets (GSE93272 and GSE17755) from the GEO database. We performed Weighted correlation network analysis (WGCNA) to analyze the expression modules in differentially expressed genes identified from GSE93272. We used KEGG, GO and GSEA enrichment analysis to elucidate the platelets-relating signatures (PRS). We then used the LASSO algorithm to develop a diagnostic model. We then used GSE17755 as a validation cohort to assess the diagnostic performance by operating Receiver Operating Curve (ROC).

Results

The application of WGCNA resulted in the identification of 11 distinct co-expression modules. Notably, Module 2 exhibited a prominent association with platelets among the differentially expressed genes (DEGs) analyzed. Furthermore, a predictive model consisting of six genes (MAPK3, ACTB, ACTG1, VAV2, PTPN6, and ACTN1) was constructed using LASSO coefficients. The resultant PRS model demonstrated excellent diagnostic accuracy in both cohorts, as evidenced by area under the curve (AUC) values of 0.801 and 0.979.

Conclusion

We elucidated the PRSs occurred in the pathogenesis of RA and developed a diagnostic model with excellent diagnostic potential.

A survey of artificial intelligence in rheumatoid arthritis

The article offers a survey of currently notable artificial intelligence methods (released between 2019-2023), with a particular emphasis on the latest advancements in detecting rheumatoid arthritis (RA) at an early stage, providing early treatment, and managing the disease. We discussed challenges in these areas followed by specific artificial intelligence (AI) techniques and summarized advances, relevant strengths, and obstacles. Overall, the application of AI in the fields of RA has the potential to enable healthcare professionals to detect RA at an earlier stage, thereby facilitating timely intervention and better disease management. However, more research is required to confirm the precision and dependability of AI in RA, and several problems such as technological and ethical concerns related to these approaches must be resolved before their widespread adoption.

Five genes as diagnostic biomarkers of dermatomyositis and their correlation with immune cell infiltration

Background

Dermatomyositis (DM) is a rare autoimmune disease characterized by severe muscle dysfunction, and the immune response of the muscles plays an important role in the development of DM. Currently, the diagnosis of DM relies on symptoms, physical examination, and biopsy techniques. Therefore, we used machine learning algorithm to screen key genes, and constructed and verified a diagnostic model composed of 5 key genes. In terms of immunity, The relationship between 5 genes and immune cell infiltration in muscle samples was analyzed. These diagnostic and immune-cell-related genes may contribute to the diagnosis and treatment of DM.

Methods

GSE5370 and GSE128470 datasets were utilised from the Gene Expression Omnibus database as DM test sets. And we also used R software to merge two datasets and to analyze the results of differentially expressed genes (DEGs) and functional correlation analysis. Then, we could detect diagnostic genes adopting least absolute shrinkage and selection operator (LASSO) logistic regression and support vector machine recursive feature elimination (SVM-RFE) analyses. The validity of putative biomarkers was assessed using the GSE1551 dataset, and we confirmed the area under the receiver operating characteristic curve (AUC) values. Finally, CIBERSORT was used to evaluate immune cell infiltration in DM muscles and the correlations between disease-related biomarkers and immune cells.

Results

In this study, a total of 414 DEGs were screened. ISG15, TNFRSF1A, GUSBP11, SERPINB1 and PTMA were identified as potential DM diagnostic biomarkers(AUC > 0.85),and the expressions of 5 genes in DM group were higher than that in healthy group (p < 0.05). Immune cell infiltration analyses indicated that identified DM diagnostic biomarkers may be associated with M1 macrophages, activated NK cells, Tfh cells, resting NK cells and Treg cells.

Conclusion

The study identified that ISG15, TNFRSF1A, GUSBP11, SERPINB1 and PTMA as potential diagnostic biomarkers of DM and these genes were closely correlated with immune cell infiltration.This will contribute to future studies in diagnosis and treatment of DM.

Screening and identification of potential hub genes and immune cell infiltration in the synovial tissue of rheumatoid arthritis by bioinformatic approach

Background

Rheumatoid arthritis (RA) is an autoimmune disease that affects individuals of all ages. The basic pathological manifestations are synovial inflammation, pannus formation, and erosion of articular cartilage, bone destruction will eventually lead to joint deformities and loss of function. However, the specific molecular mechanisms of synovitis tissue in RA are still unclear. Therefore, this study aimed to screen and explore the potential hub genes and immune cell infiltration in RA.

Methods

Three microarray datasets (GSE12021, GSE55457, and GSE55235), from the Gene Expression Omnibus (GEO) database, have been analyzed to explore the potential hub genes and immune cell infiltration in RA. First, the LIMMA package was used to screen the differentially expression genes (DEGs) after removing the batch effect. Then the clusterProfiler package was used to perform functional enrichment analyses. Second, through weighted coexpression network analysis (WGCNA), the key module was identified in the coexpression network of the gene set. Third, the protein-protein interaction (PPI) network was constructed through STRING website and the module analysis was performed using Cytoscape software. Fourth, the CIBERSORT and ssGSEA algorithm were used to analyze the immune status of RA and healthy synovial tissue, and the associations between immune cell infiltration and RA-related diagnostic biomarkers were evaluated. Fifth, we used the quantitative reverse transcription-polymerase chain reaction (qRT-PCR) to validate the expression levels of the hub genes, and ROC curve analysis of hub genes for discriminating between RA and healthy tissue. Finally, the gene-drug interaction network was constructed using DrugCentral database, and identification of drug molecules based on hub genes using the Drug Signature Database (DSigDB) by Enrichr.

Results

A total of 679 DEGs were identified, containing 270 downregulated genes and 409 upregulated genes. DEGs were primarily enriched in immune response and chemokine signaling pathways, according to functional enrichment analysis of DEGs. WGCNA explored the co-expression network of the gene set and identified key modules, the blue module was selected as the key module associated with RA. Seven hub genes are identified when PPI network and WGCNA core modules are intersected. Immune infiltration analysis using CIBERSORT and ssGSEA algorithms revealed that multiple types of immune infiltration were found to be upregulated in RA tissue compared to normal tissue. Furthermore, the levels of 7 hub genes were closely related to the relative proportions of multiple immune cells in RA. The results of the qRT-PCR demonstrated that the relative expression levels of 6 hub genes (CD27, LCK, CD2, GZMB, IL7R, and IL2RG) were up-regulated in RA synovial tissue, compared with normal tissue. Simultaneously, ROC curves indicated that the above 6 hub genes had strong biomarker potential for RA (AUC >0.8).

Conclusions

Through bioinformatics analysis and qRT-PCR experiment, our study ultimately discovered 6 hub genes (CD27, LCK, CD2, GZMB, IL7R, and IL2RG) that closely related to RA. These findings may provide valuable direction for future RA clinical diagnosis, treatment, and associated research.

Machine learning-based warning model for chronic kidney disease in individuals over 40 years old in underprivileged areas, Shanxi Province

Introduction

Chronic kidney disease (CKD) is a progressive disease with high incidence but early imperceptible symptoms. Since China's rural areas are subject to inadequate medical check-ups and single disease screening programme, it could easily translate into end-stage renal failure. This study aimed to construct an early warning model for CKD tailored to impoverished areas by employing machine learning (ML) algorithms with easily accessible parameters from ten rural areas in Shanxi Province, thereby, promoting a forward shift of treatment time and improving patients' quality of life.

Methods

From April to November 2019, CKD opportunistic screening was carried out in 10 rural areas in Shanxi Province. First, general information, physical examination data, blood and urine specimens were collected from 13,550 subjects. Afterward, feature selection of explanatory variables was performed using LASSO regression, and target datasets were balanced using the SMOTE (synthetic minority over-sampling technique) algorithm, i.e., albuminuria-to-creatinine ratio (ACR) and α1-microglobulin-to-creatinine ratio (MCR). Next, Bagging, Random Forest (RF) and eXtreme Gradient Boosting (XGBoost) were employed for classification of ACR outcomes and MCR outcomes, respectively.

Results

12,330 rural residents were included in this study, with 20 explanatory variables. The cases with increased ACR and increased MCR represented 1,587 (12.8%) and 1,456 (11.8%), respectively. After conducting LASSO, 14 and 15 explanatory variables remained in these two datasets, respectively. Bagging, RF, and XGBoost performed well in classification, with the AUC reaching 0.74, 0.87, 0.87, 0.89 for ACR outcomes and 0.75, 0.88, 0.89, 0.90 for MCR outcomes. The five variables contributing most to the classification of ACR outcomes and MCR outcomes constituted SBP, TG, TC, and Hcy, DBP and age, TG, SBP, Hcy and FPG, respectively. Overall, the machine learning algorithms could emerge as a warning model for CKD.

Conclusion

ML algorithms in conjunction with rural accessible indexes boast good performance in classification, which allows for an early warning model for CKD. This model could help achieve large-scale population screening for CKD in poverty-stricken areas and should be promoted to improve the quality of life and reduce the mortality rate.

Chemokines and chemokine receptors as promising targets in rheumatoid arthritis

Rheumatoid arthritis (RA) is an autoimmune disease that commonly causes inflammation and bone destruction in multiple joints. Inflammatory cytokines, such as IL-6 and TNF-α, play important roles in RA development and pathogenesis. Biological therapies targeting these cytokines have revolutionized RA therapy. However, approximately 50% of the patients are non-responders to these therapies. Therefore, there is an ongoing need to identify new therapeutic targets and therapies for patients with RA. In this review, we focus on the pathogenic roles of chemokines and their G-protein-coupled receptors (GPCRs) in RA. Inflamed tissues in RA, such as the synovium, highly express various chemokines to promote leukocyte migration, tightly controlled by chemokine ligand-receptor interactions. Because the inhibition of these signaling pathways results in inflammatory response regulation, chemokines and their receptors could be promising targets for RA therapy. The blockade of various chemokines and/or their receptors has yielded prospective results in preclinical trials using animal models of inflammatory arthritis. However, some of these strategies have failed in clinical trials. Nonetheless, some blockades showed promising results in early-phase clinical trials, suggesting that chemokine ligand-receptor interactions remain a promising therapeutic target for RA and other autoimmune diseases.

Identification and Validation of Hub Genes for Predicting Treatment Targets and Immune Landscape in Rheumatoid Arthritis

Background

Rheumatoid arthritis (RA) is recognized as a chronic inflammatory disease featured by pathological synovial inflammation. Currently, the underlying pathophysiological mechanisms of RA remain unclear. In the study, we attempted to explore the underlying mechanisms of RA and provide potential targets for the therapy of RA via bioinformatics analysis.

Methods

We downloaded four microarray datasets (GSE77298, GSE55235, GSE12021, and GSE55457) from the GEO database. Firstly, GSE77298 and GSE55457 were identified DEGs by the “limma” and “sva” packages of R software. Then, we performed GO, KEGG, and GSEA enrichment analyses to further analyze the function of DEGs. Hub genes were screened using LASSO analysis and SVM-RFE analysis. To further explore the differences of the expression of hub genes in healthy control and RA patient synovial tissues, we calculated the ROC curves and AUC. The expression levels of hub genes were verified in synovial tissues of normal and RA rats by qRT-PCR and western blot. Furthermore, the CIBERSORTx was implemented to assess the differences of infiltration in 22 immune cells between normal and RA synovial tissues. We explored the association between hub genes and infiltrating immune cells.

Results

CRTAM, CXCL13, and LRRC15 were identified as RA's potential hub genes by machine learning and LASSO algorithms. In addition, we verified the expression levels of three hub genes in the synovial tissue of normal and RA rats by PCR and western blot. Moreover, immune cell infiltration analysis showed that plasma cells, T follicular helper cells, M0 macrophages, M1 macrophages, and gamma delta T cells may be engaged in the development and progression of RA.

Conclusions

In brief, our study identified and validated that three hub genes CRTAM, CXCL13, and LRRC15 might involve in the pathological development of RA, which could provide novel perspectives for the diagnosis and treatment with RA.

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

Background

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.

Methods

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.

Results

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.

Conclusion

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.

Using random forest algorithm for glomerular and tubular injury diagnosis

Objectives

Chronic kidney disease (CKD) is a common chronic condition with high incidence and insidious onset. Glomerular injury (GI) and tubular injury (TI) represent early manifestations of CKD and could indicate the risk of its development. In this study, we aimed to classify GI and TI using three machine learning algorithms to promote their early diagnosis and slow the progression of CKD.

Methods

Demographic information, physical examination, blood, and morning urine samples were first collected from 13,550 subjects in 10 counties in Shanxi province for classification of GI and TI. Besides, LASSO regression was employed for feature selection of explanatory variables, and the SMOTE (synthetic minority over-sampling technique) algorithm was used to balance target datasets, i.e., GI and TI. Afterward, Random Forest (RF), Naive Bayes (NB), and logistic regression (LR) were constructed to achieve classification of GI and TI, respectively.

Results

A total of 12,330 participants enrolled in this study, with 20 explanatory variables. The number of patients with GI, and TI were 1,587 (12.8%) and 1,456 (11.8%), respectively. After feature selection by LASSO, 14 and 15 explanatory variables remained in these two datasets. Besides, after SMOTE, the number of patients and normal ones were 6,165, 6,165 for GI, and 6,165, 6,164 for TI, respectively. RF outperformed NB and LR in terms of accuracy (78.14, 80.49%), sensitivity (82.00, 84.60%), specificity (74.29, 76.09%), and AUC (0.868, 0.885) for both GI and TI; the four variables contributing most to the classification of GI and TI represented SBP, DBP, sex, age and age, SBP, FPG, and GHb, respectively.

Conclusion

RF boasts good performance in classifying GI and TI, which allows for early auxiliary diagnosis of GI and TI, thus facilitating to help alleviate the progression of CKD, and enjoying great prospects in clinical practice.

Using a Machine Learning Approach to Identify Key Biomarkers for Renal Clear Cell Carcinoma

Background

The most common and deadly subtype of renal carcinoma is kidney renal clear cell carcinoma (KIRC), which accounts for approximately 75% of renal carcinoma. However, the main cause of death in KIRC patients is tumor metastasis. There are no obvious clinical features in the early stage of kidney cancer, and 25–30% of patients have already metastasized when they are first diagnosed. Moreover, KIRC patients whose local tumors have been removed by nephrectomy are still at high risk of metastasis and recurrence and are not sensitive to chemotherapy and radiotherapy, leading to poor prognosis. Therefore, early diagnosis and treatment of this disease are very important.

Methods

KIRC-related patient datasets were downloaded from the GEO database and TCGA database. DEG screening and GO, KEGG and GSEA enrichment analysis was firstly conducted and then the LASSO and support vector machine (SVM) RFE algorithms were adopted to identify KIRC-associated key genes in training sets and validate them in the test set. The clinical prognostic analysis including the association between the expression of key genes and the overall survival, stage, grade across KIRC, the immune infiltration difference between normal samples and cancer samples, the correlation between the key genes and immune cells, immunomodulator, immune subtypes of KIRC were investigated in this research.

Results

We finally screened out 4 key genes, including ACPP, ANGPTL4, SCNN1G, SLC22A7. The expression of key genes show difference among normal samples and tumor samples, SCNN1G and SLC22A7 could be predictor of prognosis of patients. The expression of key genes was related with the abundance of tumor infiltration immune cells and the gene expression of immune checkpoint.

Conclusion

This study screened the 4 key genes, which contributed to early diagnosis, prognosis assessment and immune target treatment of patients with KIRC.

Screening of Potential Biomarkers in the Peripheral Serum for Steroid-Induced Osteonecrosis of the Femoral Head Based on WGCNA and Machine Learning Algorithms

Background. Steroid-induced osteonecrosis of the femoral head (SONFH) has produced a substantial burden of medical and social experience. However, the current diagnosis is still limited. Thus, this study is aimed at identifying potential biomarkers in the peripheral serum of patients with SONFH. Methods. The expression profile data of SONFH (number: GSE123568) was acquired from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) in SONFH were identified and used for weighted gene coexpression network analysis (WGCNA). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed to investigate the biological functions. The protein-protein interaction (PPI) network and machine learning algorithms were employed to screen for potential biomarkers. Gene set enrichment analysis (GSEA), transcription factor (TF) enrichment analysis, and competing endogenous RNA (ceRNA) network were used to determine the biological functions and regulatory mechanisms of the potential biomarkers. Results. A total of 562 DEGs, including 318 upregulated and 244 downregulated genes, were identified between SONFH and control samples, and 94 target genes were screened based on WGCNA. Moreover, biological function analysis suggested that target genes were mainly involved in erythrocyte differentiation, homeostasis and development, and myeloid cell homeostasis and development. Furthermore, GYPA, TMCC2, and BPGM were identified as potential biomarkers in the peripheral serum of patients with SONFH based on machine learning algorithms and showed good diagnostic values. GSEA revealed that GYPA, TMCC2, and BPGM were mainly involved in immune-related biological processes (BPs) and signaling pathways. Finally, we found that GYPA might be regulated by hsa-miR-3137 and that BPGM might be regulated by hsa-miR-340-3p. Conclusion. GYPA, TMCC2, and BPGM are potential biomarkers in the peripheral serum of patients with SONFH and might affect SONFH by regulating erythrocytes and immunity.

Incidence and risk factors for vertebral fracture in rheumatoid arthritis: an update meta-analysis

OBJECTIVES

This study was conducted to investigate the prevalence of vertebral fracture (VF) and its risk in patients with rheumatoid arthritis (RA) as compared to healthy individuals, and to explore the underlying risk factors.

METHODS

The electronic databases of PubMed, EMBASE, and the Cochrane Library were applied to search for the relevant literatures, which reported the prevalence of VF in both RA patients and healthy controls (up to Mar 31, 2021). The non-weighted prevalence of VF, pooled estimates of odds ratio (OR), and its 95% confidence intervals (CI) were calculated with the use of random-effects model; between-study heterogeneity was evaluated by Cochrane Q statistic, then was quantified with I². Publication bias was evaluated using Egger's linear regression test.

RESULTS

A number of 867 literatures were identified after searching for online databases, based on the inclusion and exclusion criteria, 11 eligible studies, which comprising 3805 RA patients and 59,517 healthy participants, were finally incorporated in meta-analysis. The results showed that RA patients had an increased prevalence of VF (20.29 vs 8.63%), and an elevated risk for VF (OR = 3.04, 95% CI 1.97-4.71) as compared to healthy population. Additional subgroup analysis suggested that age, body mass index (BMI), disease activity, and drug therapy might be associated with risk of VF in RA.

CONCLUSIONS

Overall, our study demonstrated an increased risk of VF in patients with RA, suggesting that age, race, BMI, disease activity, and drug therapy may be represented as risk factors contributing to the occurrence of VF in RA. Key Points • RA patients had the increased prevalence and risk of vertebral fracture (VF) as compared to healthy population. • Age, race, BMI, disease activity, and drug therapy might be associated with VF in RA. • Our findings would be helpful for the early evaluation of RA patients with high VF risk.

Screening of potential biomarkers in peripheral blood of patients with depression based on weighted gene co-expression network analysis and machine learning algorithms

BACKGROUND

The prevalence of depression has been increasing worldwide in recent years, posing a heavy burden on patients and society. However, the diagnostic and therapeutic tools available for this disease are inadequate. Therefore, this research focused on the identification of potential biomarkers in the peripheral blood of patients with depression.

METHODS

The expression dataset GSE98793 of depression was provided by the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/gds). Initially, differentially expressed genes (DEGs) were detected in GSE98793. Subsequently, the most relevant modules for depression were screened according to weighted gene co-expression network analysis (WGCNA). Finally, the identified DEGs were mapped to the WGCNA module genes to obtain the intersection genes. In addition, Gene Ontology (GO), Disease Ontology (DO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analyses were conducted on these genes. Moreover, biomarker screening was carried out by protein-protein interaction (PPI) network construction of intersection genes on the basis of various machine learning algorithms. Furthermore, the gene set enrichment analysis (GSEA), immune function analysis, transcription factor (TF) analysis, and the prediction of the regulatory mechanism were collectively performed on the identified biomarkers. In addition, we also estimated the clinical diagnostic ability of the obtained biomarkers, and performed Mfuzz expression pattern clustering and functional enrichment of the most potential biomarkers to explore their regulatory mechanisms. Finally, we also perform biomarker-related drug prediction.

RESULTS

Differential analysis was used for obtaining a total of 550 DEGs and WGCNA for obtaining 1,194 significant genes. Intersection analysis of the two yielded 140 intersection genes. Biological functional analysis indicated that these genes had a major role in inflammation-related bacterial infection pathways and cardiovascular diseases such as atherosclerosis. Subsequently, the genes S100A12, SERPINB2, TIGIT, GRB10, and LHFPL2 in peripheral serum were identified as depression biomarkers by using machine learning algorithms. Among them, S100A12 is the most valuable biomarker for clinical diagnosis. Finally, antidepressants, including disodium selenite and eplerenone, were predicted.

CONCLUSION

The genes S100A12, TIGIT, SERPINB2, GRB10, and LHFPL2 in peripheral serum are viable diagnostic biomarkers for depression. and contribute to the diagnosis and prevention of depression in clinical practice.