Screening Biomarkers for Systemic Lupus Erythematosus Based on Machine Learning and Exploring Their Expression Correlations With the Ratios of Various Immune Cells

Computer tomography-based radiomics combined with machine learning for predicting the time since onset of epidural hematoma

Estimation of the age of epidural hematoma (EDH) is a challenge in clinical forensic medicine, and this issue has yet to be conclusively resolved. The advantages of objectivity and non-invasiveness make computing tomography (CT) imaging an potential diagnostic method for EDH in living individuals. Recently, radiomics, the extraction hidden information from medical images, has emerged as a promising method for constructing predictive models. The aim of this study is to explore the feasibility and applicability of CT-based radiomics in predicting the timing of EDH injuries in surviving victims. A cohort of 95 EDH cases with definite injured time (within 12 h since injury) was selected. Clinical characteristics (age, gender, injury time, bleeding location, bleeding volume, and fracture) were recorded. The datasets were divided randomly into training and test cohorts. LIFEx software was used to segment the hematoma area in the CT and extract radiomic features. Machine learning algorithms were applied for features selection and model building. Twenty-three features were selected to calculate the Radscore, a key metric in our analysis. Utilizing this Radscore in conjunction with the time since injury, we constructed an Ordinary Least Squared (OLS) model. Our validation study has shown that mean absolute error (MAE) of the test cohort was 2.42 h, indicating a high degree of accuracy. In order to enhance the accuracy of prediction, the dataset was divided into unstable phase, occurring within the first 5 h post injury, and the stable phases. The Random Forest algorithm presented a significant divergence in predictive performance between the two phases, achieving an area under the curve (AUC) of 0.79, with an accuracy of 75.86%. The MAE of the regression model was 1.05 h for the unstable phase, and 1.23 h for the stable phase. Our findings underscore the potential of CT-based radiomics to offer a novel, convenient, and efficient approach to dating EDH, promising to illuminate new avenues in the field of medical diagnostics.

Diagnostic value of TRIM22 in diabetic kidney disease and its mechanism

PURPOSE

Diabetic kidney disease (DKD) is the primary reason of chronic kidney disease. Our objective was to discover potential autophagy-related biomarkers of tubulointerstitial injury in DKD and assess their clinical value.

METHODS

We retrieved four datasets (GSE104954, GSE30122, GSE30529, and GSE99340) of renal tubule samples from Gene Expression Omnibus (GEO) and used two algorithms (LASSO and SVM-RFE) to screen for autophagy-related differentially expressed genes (ARDEGs) in DKD. Tripartite motif containing 22 (TRIM22) was identified for subsequent validation. Validation of TRIM22 and autophagic indicators expression in clinical samples and HK-2 cells stimulated by high glucose using immunohistochemistry, immunofluorescence, and western blot.

RESULTS

We identified four ARDEGs (TRIM22, PLK2, HTR2B, and FAS) using a diagnostic gene model. ROC curves further confirmed that TRIM22 had the best diagnostic efficacy for DKD. Both clinical samples and HK-2 cells stimulated by high glucose showed high protein expression of TRIM22. The correlation analysis revealed that TRIM22 correlates with SQSTM1, NGAL, and some clinical and pathological indicators in patients with DKD.

CONCLUSION

We identified TRIM22 as a potential diagnostic biomarker for DKD, revealing its high diagnostic value in patients with DKD with moderate-to-severe interstitial fibrosis and tubular atrophy (IFTA). TRIM22 is involved in tubulointerstitial injury and autophagy dysregulation in DKD.

Gene Expression Profiling and Immune Pathway Dysregulation in Ribonucleoprotein Autoantibody-Positive Systemic Lupus Erythematosus Patients

BACKGROUND

Systemic lupus erythematosus (SLE) is an autoimmune disorder characterized by immune dysregulation and chronic inflammation across various organ systems. While anti-dsDNA and anti-Sm antibodies are commonly associated with SLE, the presence of anti-RNP antibodies is often linked to unique gene expression profiles and immune responses. This study aims to investigate the gene expression profiles in ribonucleoprotein (RNP) autoantibody-positive SLE patients by analyzing publicly available transcriptomic data.

METHODS

This study analyzed transcriptomic data from the GEO dataset GSE61635, which includes gene expression profiles from 79 anti-RNP-positive SLE patients and 30 healthy controls. Differentially expressed genes (DEGs) were identified using the GEO2R tool with a p-value < 0.05 and |log2fold change| > 1. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed. Tissue-specific and cell-type enrichment analyses highlighted the involvement of immune tissues.

RESULTS

A total of 1891 DEGs were identified between anti-RNP-positive SLE patients and healthy controls. Among the identified DEGs, SLC4A1 and EPB42 were notably downregulated, while PIP4K2A was highly upregulated. Enrichment analyses revealed significant dysregulation in antiviral response and immune regulation pathways. PPI network analysis highlighted key hub genes, suggesting a heightened antiviral state in these patients. Tissue-specific enrichment and cell-type enrichment identified the bone marrow and immune tissues as being highly affected by the altered gene expression. Additionally, gene frequency analysis highlighted RASD2 as being recurrently significant across multiple studies.

CONCLUSIONS

The findings suggest that anti-RNP-positive SLE patients exhibit distinct gene expression and immune dysregulation profiles, particularly in antiviral and immune regulation pathways. These results provide insights into the molecular mechanisms driving SLE in this patient subset and may guide future therapeutic interventions.

Machine learning and artificial intelligence within pediatric autoimmune diseases: applications, challenges, future perspective

INTRODUCTION

Autoimmune disorders affect 4.5% to 9.4% of children, significantly reducing their quality of life. The diagnosis and prognosis of autoimmune diseases are uncertain because of the variety of onset and development. Machine learning can identify clinically relevant patterns from vast amounts of data. Hence, its introduction has been beneficial in the diagnosis and management of patients.

AREAS COVERED

This narrative review was conducted through searching various electronic databases, including PubMed, Scopus, and Web of Science. This study thoroughly explores the current knowledge and identifies the remaining gaps in the applications of machine learning specifically in the context of pediatric autoimmune and related diseases.

EXPERT OPINION

Machine learning algorithms have the potential to completely change how pediatric autoimmune disorders are identified, treated, and managed. Machine learning can assist physicians in making more precise and fast judgments, identifying new biomarkers and therapeutic targets, and personalizing treatment strategies for each patient by utilizing massive datasets and powerful analytics.

Biomarkers for systemic lupus erythematosus: A scoping review

BACKGROUND

In recent years, newly discovered potential biomarkers have great research potential in the diagnosis, disease activity prediction, and treatment of systemic lupus erythematosus (SLE).

OBJECTIVE

In this study, a scoping review of potential biomarkers for SLE over several years has identified the extent to which studies on biomarkers for SLE have been conducted, the specificity, sensitivity, and diagnostic value of potential biomarkers of SLE, the research potential of these biomarkers in disease diagnosis, and activity detection is discussed.

METHODS

In PubMed and Google Scholar databases, "SLE," "biomarkers," "predictor," "autoimmune diseases," "lupus nephritis," "neuropsychiatric SLE," "diagnosis," "monitoring," and "disease activity" were used as keywords to systematically search for SLE molecular biomarkers published from 2020 to 2024. Analyze and summarize the literature that can guide the article.

CONCLUSIONS

Recent findings suggest that some potential biomarkers may have clinical application prospects. However, to date, many of these biomarkers have not been subjected to repeated clinical validation. And no single biomarker has sufficient sensitivity and specificity for SLE. It is not scientific to choose only one or several biomarkers to judge the complex disease of SLE. It may be a good direction to carry out a meta-analysis of various biomarkers to find SLE biomarkers suitable for clinical use, or to evaluate SLE by combining multiple biomarkers through mathematical models. At the same time, advanced computational methods are needed to analyze large data sets and discover new biomarkers, and strive to find biomarkers that are sensitive and specific enough to SLE and can be used in clinical practice, rather than only staying in experimental research and data analysis.

The rs3918188 and rs1799983 loci of eNOS gene are associated with susceptibility in patients with systemic lupus erythematosus in Northeast China

To investigate the association between single nucleotide polymorphism (SNP) at the rs3918188, rs1799983 and rs1007311 loci of the endothelial nitric oxide synthase (eNOS) gene and genetic susceptibility to systemic lupus erythematosus (SLE) in northeastern China. The base distribution of eNOS gene rs3918188, rs1799983 and rs1007311 in 1712 human peripheral blood samples from Northeast China was detected by SNaPshot sequencing technology. The correlation between genotype, allele and gene model of these loci of the eNOS gene and the genetic susceptibility to SLE was investigated by logistic regression analysis. The results of the differences in the frequency distribution of their gene models were visualised using R 4.3.2 software. Finally, HaploView 4.2 software was used to analyse the relationship between the haplotypes of the three loci mentioned above and the genetic susceptibility to SLE. A multifactor dimensionality reduction (MDR) analysis was used to determine the best SNP-SNP interaction model. The CC genotype and C allele at the rs3918188 locus may be a risk factor for SLE (CC vs AA: OR = 1.827, P < 0.05; C vs A: OR = 1.558, P < 0.001), and this locus increased the risk of SLE in the dominant model and the recessive model (AC + CC vs AA: OR = 1.542, P < 0.05; CC vs AA + AC: OR = 1.707, P < 0.001), while the risk of SLE was reduced in the overdominant model (AC vs AA + CC: OR = 0.628, P < 0.001). The GT genotype and T allele at locus rs1799983 may be a protective factor for SLE (GT vs GG: OR = 0.328, P < 0.001; T vs G: OR = 0.438, P < 0.001) and this locus reduced the risk of SLE in the overdominant model (GT vs GG + TT: OR = 0.385, P < 0.001). There is a strong linkage disequilibrium between the rs1007311 and rs1799983 loci of the eNOS gene. Among them, the formed haplotype AG increased the risk of SLE compared to GG. AT and GT decreased the risk of SLE compared to GG. In this study, the eNOS gene rs3918188 and rs1799983 loci were found to be associated with susceptibility to SLE. This helps to deeply explore the mechanism of eNOS gene and genetic susceptibility to SLE. It provides a certain research basis for the subsequent exploration of the molecular mechanism of these loci and SLE, as well as the early diagnosis, treatment and prognosis of SLE.

Identification of effective diagnostic genes and immune cell infiltration characteristics in small cell lung cancer by integrating bioinformatics analysis and machine learning algorithms

OBJECTIVES

To identify potential diagnostic markers for small cell lung cancer (SCLC) and investigate the correlation with immune cell infiltration.

METHODS

GSE149507 and GSE6044 were used as the training group, while GSE108055 served as validation group A and GSE73160 served as validation group B. Differentially expressed genes (DEGs) were identified and analyzed for functional enrichment. Machine learning (ML) was used to identify candidate diagnostic genes for SCLC. The area under the receiver operating characteristic curves was applied to assess diagnostic efficacy. Immune cell infiltration analyses were carried out.

RESULTS

There were 181 DEGs identified. The gene ontology analysis showed that DEGs were enriched in 455 functional annotations, some of which were associated with immunity. The kyoto encyclopedia of genes and genomes analysis revealed that there were 9 signaling pathways enriched. The disease ontology analysis indicated that DEGs were related to 116 diseases. The gene set enrichment analysis results displayed multiple items closely related to immunity. ZWINT and NRCAM were screened using ML and further validated as diagnostic genes. Significant differences were observed in SCLC with normal lung tissue samples among immune cell infiltration characteristics. Strong associations were found between the diagnostic genes and immune cell infiltration.

CONCLUSION

This study identified 2 diagnostic genes, ZWINT and NRCAM, that were related to immune cell infiltration by integrating bioinformatics analysis and ML algorithms. These genes could serve as potential diagnostic biomarkers and provide possible molecular targets for immunotherapy in SCLC.

Integrative, high-resolution analysis of single cells across experimental conditions with PARAFAC2

Effective tools for exploration and analysis are needed to extract insights from large-scale single-cell measurement data. However, current techniques for handling single-cell studies performed across experimental conditions (e.g., samples, perturbations, or patients) require restrictive assumptions, lack flexibility, or do not adequately deconvolute condition-to-condition variation from cell-to-cell variation. Here, we report that the tensor decomposition method PARAFAC2 (Pf2) enables the dimensionality reduction of single-cell data across conditions. We demonstrate these benefits across two distinct contexts of single-cell RNA-sequencing (scRNA-seq) experiments of peripheral immune cells: pharmacologic drug perturbations and systemic lupus erythematosus (SLE) patient samples. By isolating relevant gene modules across cells and conditions, Pf2 enables straightforward associations of gene variation patterns across specific patients or perturbations while connecting each coordinated change to certain cells without pre-defining cell types. The theoretical grounding of Pf2 suggests a unified framework for many modeling tasks associated with single-cell data. Thus, Pf2 provides an intuitive universal dimensionality reduction approach for multi-sample single-cell studies across diverse biological contexts.

Systemic lupus in the era of machine learning medicine

Artificial intelligence and machine learning applications are emerging as transformative technologies in medicine. With greater access to a diverse range of big datasets, researchers are turning to these powerful techniques for data analysis. Machine learning can reveal patterns and interactions between variables in large and complex datasets more accurately and efficiently than traditional statistical methods. Machine learning approaches open new possibilities for studying SLE, a multifactorial, highly heterogeneous and complex disease. Here, we discuss how machine learning methods are rapidly being integrated into the field of SLE research. Recent reports have focused on building prediction models and/or identifying novel biomarkers using both supervised and unsupervised techniques for understanding disease pathogenesis, early diagnosis and prognosis of disease. In this review, we will provide an overview of machine learning techniques to discuss current gaps, challenges and opportunities for SLE studies. External validation of most prediction models is still needed before clinical adoption. Utilisation of deep learning models, access to alternative sources of health data and increased awareness of the ethics, governance and regulations surrounding the use of artificial intelligence in medicine will help propel this exciting field forward.

HERC6 regulates STING activity in a sex-biased manner through modulation of LATS2/VGLL3 Hippo signaling

Interferon (IFN) activity exhibits a gender bias in human skin, skewed toward females. We show that HERC6, an IFN-induced E3 ubiquitin ligase, is induced in human keratinocytes through the epidermal type I IFN; IFN-κ. HERC6 knockdown in human keratinocytes results in enhanced induction of interferon-stimulated genes (ISGs) upon treatment with a double-stranded (ds) DNA STING activator cGAMP but not in response to the RNA-sensing TLR3 agonist. Keratinocytes lacking HERC6 exhibit sustained STING-TBK1 signaling following cGAMP stimulation through modulation of LATS2 and TBK1 activity, unmasking more robust ISG responses in female keratinocytes. This enhanced female-biased immune response with loss of HERC6 depends on VGLL3, a regulator of type I IFN signature. These data identify HERC6 as a previously unrecognized negative regulator of ISG expression specific to dsDNA sensing and establish it as a regulator of female-biased immune responses through modulation of STING signaling.

Machine learning approaches to identify systemic lupus erythematosus in anti-nuclear antibody-positive patients using genomic data and electronic health records

BACKGROUND

Although the 2019 EULAR/ACR classification criteria for systemic lupus erythematosus (SLE) has required at least a positive anti-nuclear antibody (ANA) titer (≥ 1:80), it remains challenging for clinicians to identify patients with SLE. This study aimed to develop a machine learning (ML) approach to assist in the detection of SLE patients using genomic data and electronic health records.

METHODS

Participants with a positive ANA (≥ 1:80) were enrolled from the Taiwan Precision Medicine Initiative cohort. The Taiwan Biobank version 2 array was used to detect single nucleotide polymorphism (SNP) data. Six ML models, Logistic Regression, Random Forest (RF), Support Vector Machine, Light Gradient Boosting Machine, Gradient Tree Boosting, and Extreme Gradient Boosting (XGB), were used to identify SLE patients. The importance of the clinical and genetic features was determined by Shapley Additive Explanation (SHAP) values. A logistic regression model was applied to identify genetic variations associated with SLE in the subset of patients with an ANA equal to or exceeding 1:640.

RESULTS

A total of 946 SLE and 1,892 non-SLE controls were included in this analysis. Among the six ML models, RF and XGB demonstrated superior performance in the differentiation of SLE from non-SLE. The leading features in the SHAP diagram were anti-double strand DNA antibodies, ANA titers, AC4 ANA pattern, polygenic risk scores, complement levels, and SNPs. Additionally, in the subgroup with a high ANA titer (≥ 1:640), six SNPs positively associated with SLE and five SNPs negatively correlated with SLE were discovered.

CONCLUSIONS

ML approaches offer the potential to assist in diagnosing SLE and uncovering novel SNPs in a group of patients with autoimmunity.

Decoding the mitochondrial connection: development and validation of biomarkers for classifying and treating systemic lupus erythematosus through bioinformatics and machine learning

BACKGROUND

Systemic lupus erythematosus (SLE) is a multifaceted autoimmune disease characterized by clinical and pathological diversity. Mitochondrial dysfunction has been identified as a critical pathogenetic factor in SLE. However, the specific molecular aspects and regulatory roles of this dysfunction in SLE are not fully understood. Our study aims to explore the molecular characteristics of mitochondria-related genes (MRGs) in SLE, with a focus on identifying reliable biomarkers for classification and therapeutic purposes.

METHODS

We sourced six SLE-related microarray datasets (GSE61635, GSE50772, GSE30153, GSE99967, GSE81622, and GSE49454) from the Gene Expression Omnibus (GEO) database. Three of these datasets (GSE61635, GSE50772, GSE30153) were integrated into a training set for differential analysis. The intersection of differentially expressed genes with MRGs yielded a set of differentially expressed MRGs (DE-MRGs). We employed machine learning algorithms-random forest (RF), support vector machine (SVM), and least absolute shrinkage and selection operator (LASSO) logistic regression-to select key hub genes. These genes' classifying potential was validated in the training set and three other validation sets (GSE99967, GSE81622, and GSE49454). Further analyses included differential expression, co-expression, protein-protein interaction (PPI), gene set enrichment analysis (GSEA), and immune infiltration, centered on these hub genes. We also constructed TF-mRNA, miRNA-mRNA, and drug-target networks based on these hub genes using the ChEA3, miRcode, and PubChem databases.

RESULTS

Our investigation identified 761 differentially expressed genes (DEGs), mainly related to viral infection, inflammatory, and immune-related signaling pathways. The interaction between these DEGs and MRGs led to the identification of 27 distinct DE-MRGs. Key among these were FAM210B, MSRB2, LYRM7, IFI27, and SCO2, designated as hub genes through machine learning analysis. Their significant role in SLE classification was confirmed in both the training and validation sets. Additional analyses included differential expression, co-expression, PPI, GSEA, immune infiltration, and the construction of TF-mRNA, miRNA-mRNA, and drug-target networks.

CONCLUSIONS

This research represents a novel exploration into the MRGs of SLE, identifying FAM210B, MSRB2, LYRM7, IFI27, and SCO2 as significant candidates for classifying and therapeutic targeting.

Exploration of biomarkers for systemic lupus erythematosus by machine-learning analysis

BACKGROUND

In recent years, research on the pathogenesis of systemic lupus erythematosus (SLE) has made great progress. However, the prognosis of the disease remains poor, and high sensitivity and accurate biomarkers are particularly important for the early diagnosis of SLE.

METHODS

SLE patient information was acquired from three Gene Expression Omnibus (GEO) databases and used for differential gene expression analysis, such as weighted gene coexpression network (WGCNA) and functional enrichment analysis. Subsequently, three algorithms, random forest (RF), support vector machine-recursive feature elimination (SVM-REF) and least absolute shrinkage and selection operation (LASSO), were used to analyze the above key genes. Furthermore, the expression levels of the final core genes in peripheral blood from SLE patients were confirmed by real-time quantitative polymerase chain reaction (RT-qPCR) assay.

RESULTS

Five key genes (ABCB1, CD247, DSC1, KIR2DL3 and MX2) were found in this study. Moreover, these key genes had good reliability and validity, which were further confirmed by clinical samples from SLE patients. The receiver operating characteristic curves (ROC) of the five genes also revealed that they had critical roles in the pathogenesis of SLE.

CONCLUSION

In summary, five key genes were obtained and validated through machine-learning analysis, offering a new perspective for the molecular mechanism and potential therapeutic targets for SLE.

An interpretable machine learning pipeline based on transcriptomics predicts phenotypes of lupus patients

Machine learning (ML) has the potential to identify subsets of patients with distinct phenotypes from gene expression data. However, phenotype prediction using ML has often relied on identifying important genes without a systems biology context. To address this, we created an interpretable ML approach based on blood transcriptomics to predict phenotype in systemic lupus erythematosus (SLE), a heterogeneous autoimmune disease. We employed a sequential grouped feature importance algorithm to assess the performance of gene sets, including immune and metabolic pathways and cell types, known to be abnormal in SLE in predicting disease activity and organ involvement. Gene sets related to interferon, tumor necrosis factor, the mitoribosome, and T cell activation were the best predictors of phenotype with excellent performance. These results suggest potential relationships between the molecular pathways identified in each model and manifestations of SLE. This ML approach to phenotype prediction can be applied to other diseases and tissues.

Cell Cultures as a Versatile Tool in the Research and Treatment of Autoimmune Connective Tissue Diseases

Cell cultures are an important part of the research and treatment of autoimmune connective tissue diseases. By culturing the various cell types involved in ACTDs, researchers are able to broaden the knowledge about these diseases that, in the near future, may lead to finding cures. Fibroblast cultures and chondrocyte cultures allow scientists to study the behavior, physiology and intracellular interactions of these cells. This helps in understanding the underlying mechanisms of ACTDs, including inflammation, immune dysregulation and tissue damage. Through the analysis of gene expression patterns, surface proteins and cytokine profiles in peripheral blood mononuclear cell cultures and endothelial cell cultures researchers can identify potential biomarkers that can help in diagnosing, monitoring disease activity and predicting patient's response to treatment. Moreover, cell culturing of mesenchymal stem cells and skin modelling in ACTD research and treatment help to evaluate the effects of potential drugs or therapeutics on specific cell types relevant to the disease. Culturing cells in 3D allows us to assess safety, efficacy and the mechanisms of action, thereby aiding in the screening of potential drug candidates and the development of novel therapies. Nowadays, personalized medicine is increasingly mentioned as a future way of dealing with complex diseases such as ACTD. By culturing cells from individual patients and studying patient-specific cells, researchers can gain insights into the unique characteristics of the patient's disease, identify personalized treatment targets, and develop tailored therapeutic strategies for better outcomes. Cell culturing can help in the evaluation of the effects of these therapies on patient-specific cell populations, as well as in predicting overall treatment response. By analyzing changes in response or behavior of patient-derived cells to a treatment, researchers can assess the response effectiveness to specific therapies, thus enabling more informed treatment decisions. This literature review was created as a form of guidance for researchers and clinicians, and it was written with the use of the NCBI database.

Screening biomarkers for Sjogren's Syndrome by computer analysis and evaluating the expression correlations with the levels of immune cells

Background

Sjögren's syndrome (SS) is a systemic autoimmune disease that affects about 0.04-0.1% of the general population. SS diagnosis depends on symptoms, clinical signs, autoimmune serology, and even invasive histopathological examination. This study explored biomarkers for SS diagnosis.

Methods

We downloaded three datasets of SS patients' and healthy pepole's whole blood (GSE51092, GSE66795, and GSE140161) from the Gene Expression Omnibus (GEO) database. We used machine learning algorithm to mine possible diagnostic biomarkers for SS patients. Additionally, we assessed the biomarkers' diagnostic value using the receiver operating characteristic (ROC) curve. Moreover, we confirmed the expression of the biomarkers through the reverse transcription quantitative polymerase chain reaction (RT-qPCR) using our own Chinese cohort. Eventually, the proportions of 22 immune cells in SS patients were calculated by CIBERSORT, and connections between the expression of the biomarkers and immune cell ratios were studied.

Results

We obtained 43 DEGs that were mainly involved in immune-related pathways. Next, 11 candidate biomarkers were selected and validated by the validation cohort data set. Besides, the area under curves (AUC) of XAF1, STAT1, IFI27, HES4, TTC21A, and OTOF in the discovery and validation datasets were 0.903 and 0.877, respectively. Subsequently, eight genes, including HES4, IFI27, LY6E, OTOF, STAT1, TTC21A, XAF1, and ZCCHC2, were selected as prospective biomarkers and verified by RT-qPCR. Finally, we revealed the most relevant immune cells with the expression of HES4, IFI27, LY6E, OTOF, TTC21A, XAF1, and ZCCHC2.

Conclusion

In this paper, we identified seven key biomarkers that have potential value for diagnosing Chinese SS patients.

Application of Machine Learning Models in Systemic Lupus Erythematosus

Systemic Lupus Erythematosus (SLE) is a systemic autoimmune disease and is extremely heterogeneous in terms of immunological features and clinical manifestations. This complexity could result in a delay in the diagnosis and treatment introduction, with impacts on long-term outcomes. In this view, the application of innovative tools, such as machine learning models (MLMs), could be useful. Thus, the purpose of the present review is to provide the reader with information about the possible application of artificial intelligence in SLE patients from a medical perspective. To summarize, several studies have applied MLMs in large cohorts in different disease-related fields. In particular, the majority of studies focused on diagnosis and pathogenesis, disease-related manifestations, in particular Lupus Nephritis, outcomes and treatment. Nonetheless, some studies focused on peculiar features, such as pregnancy and quality of life. The review of published data demonstrated the proposal of several models with good performance, suggesting the possible application of MLMs in the SLE scenario.

Alterations of the gut virome in patients with systemic lupus erythematosus

Background

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease that has been linked to the dysbiosis of the gut microbiome and virome. However, the potential characterization of the gut virome in SLE patients needs to be explored more extensively.

Methods

Herein, we analyzed the gut viral community of 16 SLE patients and 31 healthy controls using both bulk and virus-like particle (VLP)-based metagenomic sequencing of their fecal samples. A total of 15,999 non-redundant viral operational taxonomic units (vOTUs) were identified from the metagenomic assembled contigs and used for gut virome profiling.

Results

SLE patients exhibited a significant decrease in gut viral diversity in the bulk metagenome dataset, but this change was not significant in the VLP metagenome dataset. Also, considerable alterations of the overall gut virome composition and remarkable changes in the viral family compositions were observed in SLE patients compared with healthy controls, as observed in both two technologies. We identified 408 vOTUs (177 SLE-enriched and 231 control-enriched) with significantly different relative abundances between patients and controls in the bulk virome, and 18 vOTUs (17 SLE-enriched in 1 control-enriched) in the VLP virome. The SLE-enriched vOTUs included numerous Siphoviridae, Microviridae, and crAss-like viruses and were frequently predicted to infect Bacteroides, Parabacteroides, and Ruminococcus_E, while the control-enriched contained numerous members of Siphoviridae and Myoviridae and were predicted to infect Prevotella and Lachnospirales_CAG-274. We explored the correlations between gut viruses and bacteria and found that some Lachnospirales_CAG-274 and Hungatella_A phages may play key roles in the virus-bacterium network. Furthermore, we explored the gut viral signatures for disease discrimination and achieved an area under the receiver operator characteristic curve (AUC) of above 0.95, suggesting the potential of the gut virome in the prediction of SLE.

Conclusion

Our findings demonstrated the alterations in viral diversity and taxonomic composition of the gut virome of SLE patients. Further research into the etiology of SLE and the gut viral community will open up new avenues for treating and preventing SLE and other autoimmune diseases.