BACKGROUND: Our study aimed to identify potential specific biomarkers for osteoarthritis (OA) and assess their relationship with immune infiltration.
METHODS: We utilized data from GSE117999, GSE51588, and GSE57218 as training sets, while GSE114007 served as a validation set, all obtained from the GEO database. First, weighted gene co-expression network analysis (WGCNA) and functional enrichment analysis were performed to identify hub modules and potential functions of genes. We subsequently screened for potential OA biomarkers within the differentially expressed genes (DEGs) of the hub module using machine learning methods. The diagnostic accuracy of the candidate genes was validated. Additionally, single gene analysis and ssGSEA was performed. Then, we explored the relationship between biomarkers and immune cells. Lastly, we employed RT-PCR to validate our results.
RESULTS: WGCNA results suggested that the blue module was the most associated with OA and was functionally associated with extracellular matrix (ECM)-related terms. Our analysis identified ALB, HTRA1, DPT, MXRA5, CILP, MPO, and PLAT as potential biomarkers. Notably, HTRA1, DPT, and MXRA5 consistently exhibited increased expression in OA across both training and validation cohorts, demonstrating robust diagnostic potential. The ssGSEA results revealed that abnormal infiltration of DCs, NK cells, Tfh, Th2, and Treg cells might contribute to OA progression. HTRA1, DPT, and MXRA5 showed significant correlation with immune cell infiltration. The RT-PCR results also confirmed these findings.
CONCLUSIONS: HTRA1, DPT, and MXRA5 are promising biomarkers for OA. Their overexpression strongly correlates with OA progression and immune cell infiltration.

BACKGROUND: Preeclampsia is a severe obstetric disorder that significantly affects the maternal and neonatal peri-partum safety and long-term quality of life. However, there is limited research exploring the common mechanisms and potential clinical significance between early-onset preeclampsia and full-term preeclampsia from an immunological perspective.
METHODS: In this study, data analysis was conducted. Initially, immune-related co-expressed genes involving both subtypes of preeclampsia were identified through Weighted Gene Co-expression Network Analysis (WGCNA). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were further employed to investigate the shared pathways regulated by immune-related genes. Binary logistic regression identified co-expressed genes with diagnostic value for preeclampsia, and a diagnostic model was constructed. Gene Set Enrichment Analysis (GSEA) predicted the potential biological functions of the selected genes. Lasso and Cox regression analyses identified genes closely associated with gestational duration, and a risk score model was established. A 4-gene feature, immune-related gene model for predicting the risk of preterm birth in preeclamptic pregnant women, was developed and validated through qPCR experiments. Immune cell infiltration analysis determined differences in immune cell infiltration between the two subtypes of preeclampsia.
RESULTS: This study identified 4 immune-related co-expressed genes (CXCR6, PIK3CB, IL1RAP, and OSMR). Additionally, diagnostic and preterm birth risk prediction models for preeclampsia were constructed based on these genes. GSEA analysis suggested the involvement of these genes in the regulation of galactose metabolism, notch signaling pathway, and RIG-I like receptor signaling pathway. Immune pathway analysis indicated that the activation of T cell co-inhibition could be a potential intervention target for immunotherapy in early-onset preeclampsia.
CONCLUSIONS: Our study provides promising insights into immunotherapy and mechanistic research for preeclampsia, discovering novel diagnostic and intervention biomarkers, and offering personalized diagnostic tools for preeclampsia.

BACKGROUND: Endoplasmic reticulum stress (ERS) could be a strategy for treating malignant tumors. Moreover, long noncoding RNAs (lncRNAs) can promote tumorigenesis and progression, and forecast the prognosis of cancers. Nevertheless, the prognostic value of ERS-related lncRNAs has not been reported in lung adenocarcinoma (LUAD).
METHODS: The messenger RNA (mRNA), microRNA (miRNA) and lncRNA expression data related to LUAD were obtained in public databases (TCGA and GEO databases). Prognostic ERS-related differentially expressed lncRNAs (ERS-DELs) were obtained and used to build an ERS-related model by Cox regression analysis. Moreover, we further screened independent prognostic elements and built a nomogram. Furthermore, enrichment analysis of genes was conducted to investigate the functions. A lncRNA-miRNA-mRNA network was built to explore mechanism of lncRNAs. Finally, qRT-PCR was utilized to examine the expression levels of lncRNAs.
RESULTS: 30 ERS-DELs were identified, and an ERS-related signature was built based on AF131215.2, LINC00472, LINC01352, RP1-78O14.1, RP11-253E3.3, RP11-98D18.9, and SNHG12. Gene set enrichment analysis indicated that genes in the high-risk group were chiefly focused on the regulation of mRNA binding, and genes in the low-risk group were significantly focused on protein localization to cilia. A lncRNA-miRNA-mRNA network, containing 7 signature lncRNAs, 23 miRNAs, and 128 mRNAs, was also established. Eventually, quantitative real-time polymerase chain reaction was used to confirm that seven prognostic lncRNAs had a consistent expression with the analysis.
CONCLUSIONS: An ERS-related signature containing seven prognostic lncRNAs was built, which offered new thinking concerning the role of ERS-related lncRNAs in LUAD.

BACKGROUND: Inflammatory cytokines such as Interleukin 1β(IL1β), IL6,Tumor Necrosis Factor-α (TNF-α) can inhibit osteoblast differentiation and induce osteoblast apoptosis. PANoptosis, a newly identified type of programmed cell death (PCD), may be influenced by long noncoding RNA (lncRNAs) which play important roles in regulating inflammation. However, the potential role of lncRNAs in inflammation and PANoptosis during osteogenic differentiation remains unclear. This study aimed to investigate the regulatory functions of lncRNAs in inflammation and apoptosis during osteogenic differentiation.
RESULTS: High-throughput sequencing was used to identify differentially expressed genes involved in osteoblast differentiation under inflammatory conditions. Two lncRNAs associated with inflammation and PANoptosis during osteogenic differentiation were identified from sequencing data and Gene Expression Omnibus (GEO) databases. Their functionalities were analyzed using diverse bioinformatics methodologies, resulting in the construction of the lncRNA-miRNA-mRNA network. Among these, lncRNA (MIR17HG) showed a high correlation with PANoptosis. Bibliometric methods were employed to collect literature data on PANoptosis, and its components were inferred. PCR and Western Blotting experiments confirmed that lncRNA MIR17HG is related to PANoptosis in osteoblasts during inflammation.
CONCLUSIONS: Our data suggest that TNF-α-induced inhibition of osteogenic differentiation and PANoptosis in MC3T3-E1 osteoblasts is associated with MIR17HG. These findings highlight the critical role of MIR17HG in the interplay between inflammation, PANoptosis, and osteogenic differentiation, suggesting potential therapeutic targets for conditions involving impaired bone formation and inflammatory responses.

UNASSIGNED: Type 2 diabetes mellitus (T2DM) is a major cause of atherosclerosis (AS). However, definitive evidence regarding the common molecular mechanisms underlying these two diseases are lacking. This study aimed to investigate the mechanisms underlying the association between T2DM and AS.
UNASSIGNED: The gene expression profiles of T2DM (GSE159984) and AS (GSE100927) were obtained from the Gene Expression Omnibus, after which overlapping differentially expressed gene identification, bioinformatics enrichment analyses, protein-protein interaction network construction, and core genes identification were performed. We confirmed the discriminatory capacity of core genes using receiver operating curve analysis. We further identified transcription factors using TRRUST database to build a transcription factor-mRNA regulatory network. Finally, the immune infiltration and the correlation between core genes and differential infiltrating immune cells were analyzed.
UNASSIGNED: A total of 27 overlapping differentially expressed genes were identified under the two-stress conditions. Functional analyses revealed that immune responses and transcriptional regulation may be involved in the potential pathogenesis. After protein-protein interaction network deconstruction, external datasets, and qRT-PCR experimental validation, four core genes (IL1B, C1QA, CCR5, and MSR1) were identified. ROC analysis further showed the reliable value of these core genes. Four common differential infiltrating immune cells (B cells, CD4+ T cells, regulatory T cells, and M2 macrophages) between T2DM and AS datasets were selected based on immune cell infiltration. A significant correlation between core genes and common differential immune cells. Additionally, five transcription factors (RELA, NFκB1, JUN, YY1, and SPI1) regulating the transcription of core genes were mined using upstream gene regulator analysis.
UNASSIGNED: In this study, common target genes and co-immune infiltration landscapes were identified between T2DM and AS. The relationship among five transcription factors, four core genes, and four immune cells profiles may be crucial to understanding T2DM complicated with AS pathogenesis and therapeutic direction.

This paper focuses on inferring a general class of hidden Markov models (HMMs) using data acquired from experts. Expert-acquired data contain decisions/actions made by humans/users for various objectives, such as navigation data reflecting drivers\' behavior, cybersecurity data carrying defender decisions, and biological data containing the biologist\'s actions (e.g., interventions and experiments). Conventional inference methods rely on temporal changes in data without accounting for expert knowledge. This paper incorporates expert knowledge into the inference of HMMs by modeling expert behavior as an imperfect reinforcement learning agent. The proposed method optimally quantifies experts\' perceptions about the system model, which, alongside the temporal changes in data, contributes to the inference process. The proposed inference method is derived through a combination of dynamic programming and optimal recursive Bayesian estimation. The applicability of this method is demonstrated to a wide range of inference criteria, such as maximum likelihood and maximum a posteriori. The performance of the proposed method is investigated through a comprehensive numerical experiment using a benchmark problem and biological networks.

Objective: To construct a repetitive implantation failure (RIF)-related competitive endogenous RNA (ceRNA) regulatory network and validate with clinical samples. Methods: RIF-related long non-coding RNA (lncRNA), microRNA (miRNA) and messenger RNA (mRNA) from the high-throughput gene expression omnibus (GEO) database Expression profile data set were obtained to construct a ceRNA regulatory network of lncRNA-miRNA-mRNA. At the same time, weighted gene co-expression network analysis (WGCNA) was used to explore hub genes in the network. This retrospective study collected RIF patients and controls (at least one pregnancy history after assisted conception) who underwent in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) for assisted pregnancy from 2020 to 2021 at the Reproductive Medicine Center of the First Affiliated Hospital of Zhengzhou University. In the endometrial tissue of patients with 1 pregnancy history, real-time fluorescence quantitative polymerase chain reaction (qRT-PCR) was used to verify the mRNA expression levels of RIF-related hub genes, and Western blotting and immunohistochemistry were used to verify protein expression levels of vascular cell adhesion molecule-1 (VCAM1). Results: A RIF-related ceRNA regulatory network consisting of 32 lncRNAs, 31 miRNAs and 88 mRNAs was constructed, and 7 RIF-related hub genes were identified using WGCNA. By intersecting 88 mRNAs and hub genes in the ceRNA network, two RIF-related key genes were obtained, i.e., VCAM1 and interleukin-2 receptor α (interleukin-2 receptor α, IL-2RA). In clinical verification, the ages of the control group and RIF group [M (Q1, Q3)] were 26.50 (25.00, 34.00) and 30.50 (25.75, 35.25) years old, respectively (P>0.05). Compared with the control group, the mRNA [0.30 (0.15, 0.42) vs 0.99 (0.69, 1.34), P=0.001] and protein expression [0.44 (0.16, 1.27) vs 2.39 (1.58, 2.58), P<0.001] of VCAM1 in the endometrium of the RIF group were both reduced. Conclusions: This study uses bioinformatics analysis methods to construct a RIF-related ceRNA regulatory network, and it is confirmed through clinical samples that the expression level of VCAM1 in the endometrial tissue of RIF patients is significantly reduced.
目的： 构建反复种植失败（RIF）相关竞争性内源RNA（ceRNA）调控网络并进行临床样本验证。 方法： 从高通量基因表达数据库（GEO）得到RIF相关的长链非编码RNA（lncRNA）、微小RNA（miRNA）和信使RNA（mRNA）表达谱数据集，构建lncRNA-miRNA-mRNA的ceRNA调控网络。同时利用加权基因共表达网络分析（WGCNA）探索网络中的枢纽基因。回顾性收集2020—2021年于郑州大学第一附属医院生殖医学中心行体外受精（IVF）/卵胞浆内单精子显微注射（ICSI）助孕的RIF患者和对照组（助孕后至少有1次妊娠史）患者的子宫内膜组织，应用实时荧光定量聚合酶链反应（qRT-PCR）验证RIF相关枢纽基因的mRNA表达水平，并应用Western印迹和免疫组化技术验证血管细胞黏附分子-1（VCAM1）的蛋白表达水平。 结果： 构建了由32个lncRNA、31个miRNA和88个mRNA组成的RIF相关ceRNA调控网络，并利用WGCNA鉴定出7个RIF相关枢纽基因。将ceRNA网络中的88个mRNA与枢纽基因取交集得到2个RIF相关关键基因：VCAM1和白细胞介素2受体α（IL-2RA）。临床验证中，对照组和RIF组的年龄［M（Q1，Q3）］分别为26.50（25.00，34.00）和30.50（25.75，35.25）岁（P>0.05）。与对照组相比，RIF组子宫内膜中VCAM1的mRNA［M（Q1，Q3）］［0.30（0.15，0.42）比0.99（0.69，1.34），P=0.001］和蛋白表达水平［M（Q1，Q3）］［0.44（0.16，1.27）比2.39（1.58，2.58），P<0.001］均降低。 结论： 本研究成功构建了RIF相关ceRNA调控网络，并通过临床样本证实RIF患者子宫内膜组织中VCAM1的表达水平降低。.

BACKGROUND: Changes in the oxidative stress and lipid metabolism (OSLM) pathways play important roles in polycystic ovarian syndrome (PCOS) pathogenesis and development. Consequently, a systematic analysis of genes related to OSLM was conducted to identify molecular clusters and explore new biomarkers that are helpful for the diagnostic of PCOS.
METHODS: Gene expression and clinical data from 22 PCOS women and 14 normal women were obtained from the GEO database (GSE34526, GSE95728, and GSE106724). Consensus clustering identified OSLM-related molecular clusters, and WGCNA revealed co-expression patterns. The immune microenvironment was quantitatively assessed utilizing the CIBERSORT algorithm. Multiple machine learning models and connectivity map analyses were subsequently applied to explore potential biomarkers for PCOS, and nomograms were employed to develop a predictive multigene model of PCOS. Finally, the OSLM status of PCOS and the hub genes expression profiles were preliminarily verified using TUNEL, qRT‒PCR, western blot, and IHC assays in a PCOS mouse model.
RESULTS: 19 differential expression genes (DEGs) related to OSLM were identified. Based on 19 DEGs that were strongly influenced by OSLM, PCOS patients were stratified into two distinct clusters, designated Cluster 1 and Cluster 2. Distinct differences in the immune cell proportions existed in normal and two PCOS clusters. The random forest showed the best results, with the least cross-entropy and the utmost AUC (cross-entropy: 0.111 AUC: 0.960). Among the 19 OSLM-related genes, CXCR1, ACP5, CEACAM3, S1PR4, and TCF7 were identified by a Bayesian network and had a good fit with PCOS disease risk by the nomogram (AUC: 0.990 CI: 0.968-1.000). TUNEL assays revealed more severe DNA damage within the ovarian granule cells of PCOS mice than in those of normal mice (P < 0.001). The RNA and protein expression levels of the five hub genes were significantly elevated in PCOS mice, which was consistent with the results of the bioinformatics analyses.
CONCLUSIONS: A novel predictive model was constructed for PCOS patients and five hub genes were identified as potential biomarkers to offer novel insights into clinical diagnostic strategies for PCOS.

BACKGROUND: Activation of brown adipose tissue (BAT) has gained attention due to its ability to dissipate energy and counteract cardiometabolic diseases (CMDs).
METHODS: This study investigated the consequences of cold exposure on the BAT and liver proteomes of an established CMD mouse model based on LDL receptor-deficient (LdlrKO) mice fed a high-fat, high-sucrose, high-cholesterol diet for 16 weeks. We analyzed energy metabolism in vivo and performed untargeted proteomics on BAT and liver of LdlrKO mice maintained at 22 °C or 5 °C for 7 days.
RESULTS: We identified several dysregulated pathways, miRNAs, and transcription factors in BAT and liver of cold-exposed Ldlrko mice that have not been previously described in this context. Networks of regulatory interactions based on shared downstream targets and analysis of ligand-receptor pairs identified fibrinogen alpha chain (FGA) and fibronectin 1 (FN1) as potential crosstalk factors between BAT and liver in response to cold exposure. Importantly, genetic variations in the genes encoding FGA and FN1 have been associated with cardiometabolic-related phenotypes and traits in humans.
CONCLUSIONS: This study describes the key factors, pathways, and regulatory networks involved in the crosstalk between BAT and the liver in a cold-exposed CMD mouse model. These findings may provide a basis for future studies aimed at testing whether molecular mediators, as well as regulatory and signaling mechanisms involved in tissue adaption upon cold exposure, could represent a target in cardiometabolic disorders.

BACKGROUND: Increasing evidence suggests that a substantial proportion of disease-associated mutations occur in enhancers, regions of non-coding DNA essential to gene regulation. Understanding the structures and mechanisms of the regulatory programs this variation affects can shed light on the apparatuses of human diseases.
RESULTS: We collect epigenetic and gene expression datasets from seven early time points during neural differentiation. Focusing on this model system, we construct networks of enhancer-promoter interactions, each at an individual stage of neural induction. These networks serve as the base for a rich series of analyses, through which we demonstrate their temporal dynamics and enrichment for various disease-associated variants. We apply the Girvan-Newman clustering algorithm to these networks to reveal biologically relevant substructures of regulation. Additionally, we demonstrate methods to validate predicted enhancer-promoter interactions using transcription factor overexpression and massively parallel reporter assays.
CONCLUSIONS: Our findings suggest a generalizable framework for exploring gene regulatory programs and their dynamics across developmental processes; this includes a comprehensive approach to studying the effects of disease-associated variation on transcriptional networks. The techniques applied to our networks have been published alongside our findings as a computational tool, E-P-INAnalyzer. Our procedure can be utilized across different cellular contexts and disorders.