Xia Xu

Xia Xu


Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 May 2023 in BMC Pregnancy and Childbirth
Verified
Journal Article
Open access
PDF file
TLTingting Liao
Tingting Liao
XXXia Xu
Xia Xu
YZYulong Zhang
Yulong Zhang
YJYan Jianying
Yan Jianying
Objective: To study the combined effect of gestational diabetes mellitus (GDM) and maximum level of maternal serum total bile acid (TBA) on the incidence of adverse pregnancy outcomes in women with intrahepatic cholestasis of pregnancy (ICP). Methods: This was an observational study with 724 women with ICP. Perinatal outcomes were compared by the presence of GDM. Logistic regression was used to assess the independent and multiplicative interactions of GDM and maximum maternal serum TBA on adverse pregnancy outcomes. Additive interactions were calculated using an Excel sheet developed by Andersson to calculate relative excess risks. Results: The incidence of GDM in patients with ICP was 21.55%. Maternal age, pre-pregnancy weight, parity, and gravidity were positively correlated with GDM. Hypertensive disorders of pregnancy (HDP) and fetal distress rates were higher in the GDM vs. non-GDM group. There were no significant differences in biochemical outcomes (i.e., Triglyceride (TG), low density lipoprotein (LDL), alanine aminotransferase (ALT), aspartate aminotransferase (AST) and total bile acid (TBA)) between the two groups. In terms of adverse pregnancy outcomes, GDM was only associated with maximum TBA concentration for cesarean section. No additive or pairwise interactions were detected between GDM and maximum TBA concentration and HDP, PPH, preterm delivery, LGA, SGA, and cesarean section. Conclusion: GDM independently contributes to adverse pregnancy outcomes among women with ICP. However, the combined effects of GDM and maximum TBA concentration on adverse pregnancy outcomes do not appear to be multiplicative or additive.

12 March 2025 in Frontiers in Genetics
Verified
Journal Article
Open access
PDF file
TYTiantian Yu
Tiantian Yu
GWGuiying Wang
Guiying Wang
XXXia Xu
Xia Xu
YJYan Jianying
Yan Jianying
Background: Preeclampsia (PE), a major obstetric disorder marked by dysfunction in both placental and maternal vascular systems, continues to pose critical challenges in global maternal healthcare. This multisystem pregnancy complication contributes significantly to adverse perinatal outcomes and remains a leading cause of pregnancy-related morbidity worldwide. However, the available treatment options at present remain restricted. Our investigation employs an integrative bioinformatics approach to elucidate critical molecular signatures linked to the interplay between immunological dysregulation and oxidative stress mechanisms in PE pathogenesis.Methods: In this study, we sourced the dataset from the GEO database with the aim of pinpointing differentially expressed genes (DEGs) between PE samples and control samples. Genes associated with oxidative stress were procured from the Genecards database. Next, we employed a comprehensive approach. This involved integrating WGCNA, GO and KEGG pathway analyses, constructing PPI networks, applying machine learning algorithms, performing gene GSEA, and conducting immune infiltration analysis to identify the key hub genes related to oxidative stress. Diagnostic potential of candidate biomarkers was quantitatively assessed through ROC curve modeling. Additionally, we constructed a miRNA - gene regulatory network for the identified diagnostic genes and predicted potential candidate drugs. In the final step, we validated the significant hub gene using independent external datasets, the hypoxia model of the HTR-8/SVneo cell line, and human placental tissue samples.Results: At last, leptin (LEP) was identified as a core gene through screening and was found to be upregulated. The results of quantitative real-time polymerase chain reaction (qRT -PCR) and immunohistochemistry validation were consistent with those obtained from the datasets. KEGG analysis revealed that LEP was significantly enriched in “allograft rejection,” “antigen processing,” “ECM receptor interaction” and “graft versus host disease.” GO analysis revealed that LEP was involved in biological processes such as “antigen processing and presentation,” “peptide antigen assembly with MHC protein complex,” “complex of collagen trimers,” “MHC class II protein complex” and “mitochondrial protein containing complex.” Moreover, immune cell analysis indicated that T follicular helper cells, plasmacytoid dendritic cells, neutrophils, and activated dendritic cells were positively correlated with LEP expression, whereas γδT cells, eosinophils, and central memory CD4+ T cells showed a negative correlation. These findings suggest that LEP influences the immune microenvironment of PE through its interaction with arious immune cells. In addition, 28 miRNAs and 15 drugs were predicted to target LEP. Finally, the overexpression of LEP was verified using independent external datasets, the hypoxia model of the HTR-8/SVneo cell line, and human placental tissue.Conclusion: Through an integrated analytical framework employing WGCNA coupled with three distinct machine learning-driven phenotypic classification models, we discovered a pivotal regulatory gene. This gene has the potential to act as a novel diagnostic biomarker for PE. Moreover, it can be considered as a promising target for drug development related to PE. Notably, it shows a strong correlation with the immune microenvironment, suggesting its crucial role in the complex pathophysiological processes underlying PE.

1 August 2025 in International Journal of General Medicine
Verified
Journal Article
Open access
PDF file
YXYanhong Xu
Yanhong Xu
YZYizheng Zu
Yizheng Zu
YZYing Zhang
Ying Zhang
ZLZewei Liang
Zewei Liang
XXXia Xu
Xia Xu
YJYan Jianying
Yan Jianying
Purpose: To analyze the risk factors for preterm birth in patients with early-onset preeclampsia (EOPE) based on multi-algorithm machine learning and to construct a predictive model to explore the predictive value of the model. Methods: A retrospective analysis was conducted on 442 EOPE patients from a single tertiary center, divided into preterm birth (< 37 weeks, n=358) and term-born (≥ 37 weeks, n=84) groups. Univariate analysis, random forest importance assessment, lasso regression combined with multivariate regression analysis were used for feature evaluation. Eight machine learning models were trained (70% data) and validated (30% data). A Stacking ensemble model was constructed, and SHapley Additive exPlanations (SHAP) was used for feature interpretation. Results: The area under the receiver operating characteristic curve (AUROC) for predicting preterm birth in EOPE patients using Logistic Regression, Gaussian Naive Bayes, Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine, Support Vector Machine (SVM), Gradient Boosting Decision Tree (GBDT), Multi-Layer Perceptron, and Elastic Net were 0.763, 0.712, 0.821, 0.832, 0.821, 0.842, 0.784, and 0.763, respectively. The Stacking model (XGBoost+GBDT+SVM) achieved superior performance (AUROC=0.865). Three independent risk factors were identified: fetal growth restriction (aOR=3.50, p = 0.047), serum cystatin C (aOR=11.27, p = 0.018), and C-reactive protein (aOR=1.37, p < 0.001). SHAP analysis revealed GBDT as the top contributor to Stacking predictions, with microalbunminuria (GBDT, XGBoost) and age (SVM) being the most influential features. Conclusion: Machine learning models can serve as reliable assessment tools for predicting the risk of preterm birth in patients with EOPE. The ensemble prediction model demonstrates the best predictive performance, helping obstetricians identify high-risk patients and perform early intervention to improve perinatal outcomes.

24 July 2025 in BMC Pregnancy and Childbirth
Verified
Journal Article
Open access
PDF file
YXYanhong Xu
Yanhong Xu
YZYizheng Zu
Yizheng Zu
XLXiaosi Lu
Xiaosi Lu
YWYiping Wang
Yiping Wang
JZJiaying Zheng
Jiaying Zheng
YJYan Jianying
Yan Jianying
Preeclampsia (PE) is a complex disorder with significant variability in organ involvement. It remains unclear whether machine learning can identify organ-system-based subclasses of PE. This study aimed to identify novel subclasses of PE using organ-system biomarkers to improve pregnancy outcome prediction. We retrospectively analyzed clinical data from PE patients at Fujian Maternity and Child Health Hospital. K-means clustering was applied using organ system function indicators, with 10-fold cross-validation. Functional indicators and pregnancy outcomes were compared across subclasses. Heatmap and sankey diagrams were used to reveal the distribution of patients across combined organ system clusters. The analysis included 7,531 PE patients treated between 2013 and 2023. 10-fold cross-validation confirmed clustering robustness with mean ARI of 0.8806 ± 0.0099 and NMI of 0.7800 ± 0.0123. Three heart function clusters were identified using five indicators, with H-Cluster 1 showing the poorest heart function and the highest complication rates. Five kidney clusters were determined from ten indicators. K-Cluster 1 and K-Cluster 5 showed distinct biomarker patterns but similar complication rates (P > 0.05). Liver function analysis using thirteen indicators revealed four clusters. L-Cluster 1 exhibited elevated liver enzymes and bilirubin with higher severe PE and intrahepatic cholestasis rates, whereas L-Cluster 3 had lower protein levels but higher anemia, fetal distress and hemorrhage incidence (P < 0.05). Five coagulation clusters were determined from nine indicators, showing significant differences in indicators and complication rates (P < 0.05). Heatmap and sankey diagram analyses revealed significant overlap between high-risk clusters, with the most frequent combination being H-Cluster 1, K-Cluster 1, L-Cluster 1 and C-Cluster 5. Machine learning identified distinct PE subclasses based on organ system dysfunction patterns, each demonstrating unique pregnancy outcomes. This suggests potential clinical utility of computational approaches for PE subclassification and generates hypotheses for further investigation of its biological mechanisms.

24 January 2025 in The Journal of Clinical Hypertension
Verified
Journal Article
Open access
TYTiantian Yu
Tiantian Yu
GWGuiying Wang
Guiying Wang
XXXia Xu
Xia Xu
YJYan Jianying
Yan Jianying
Preeclampsia (PE) is a pregnancy‐specific disorder characterized by an unclearly understood pathogenesis and poses a great threat to maternal and fetal safety. Cuproptosis, a novel form of cellular death, has been implicated in the advancement of various diseases. However, the role of cuproptosis and immune‐related genes in PE is unclear. The current study aims to elucidate the gene expression matrix and immune infiltration patterns of cuproptosis‐related genes (CRGs) in the context of PE. The GSE98224 dataset was obtained from the Gene Expression Omnibus (GEO) database and utilized as the internal training set. Based on the GSE98224 dataset, we explored the differentially expressed cuproptosis related genes (DECRGs) and immunological composition. We identified 10 DECRGs conducted Gene Ontology (GO) function, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses, and a protein–protein interaction (PPI) network. Furthermore, patients with PE were categorized into two distinct clusters, and an investigation was conducted to examine the status of immune cell infiltration. Additionally, the application of Weighted Gene Co‐expression Network Analysis (WGCNA) was utilized to differentiate modules consisting of co‐expressed genes and conduct clustering analysis. The intersecting genes were obtained by intersecting differently expressed genes in PE and PE clusters. The most precise forecasting model was chosen by evaluating the effectiveness of four machine learning models. The ResNet model was established to score the hub genes. The prediction accuracy was assessed by receiver operating characteristic (ROC) curves and an external dataset. We successfully identified five key DECREGs and two pathological clusters in PE, each with distinct immune profiles and biological characteristics. Subsequently, the RF model was deemed the most optimal model for the identification of PE with a large area under the curve (AUC = 0.733). The five genes that ranked highest in the RF machine learning model were considered to be predictor genes. The calibration curve demonstrated a high level of accuracy in aligning the predicted outcomes with the actual outcomes. We validate the ResNet model using the ROC curve with the area under the curve (AUC = 0.82). Cuproptosis and immune infiltration may play an important role in the pathogenesis of PE. The present study elucidated that GSTA4, KCNK5, APLNR, IKZF2, and CAP2 may be potential markers of cuproptosis‐associated PE and are considered to play a significant role in the initiation and development of cuproptosis‐induced PE.

6 November 2025 in BMC Medical Informatics and Decision Making
Verified
Journal Article
Open access
PDF file
YXYanhong Xu
Yanhong Xu
XLXinying Liu
Xinying Liu
YZYing Zhang
Ying Zhang
XQXingyi Qi
Xingyi Qi
CJChengcheng Jin
Chengcheng Jin
JYJianying Yan
Jianying Yan
Early-onset preeclampsia (PE) poses significant risks for maternal and fetal outcomes, particularly when emergency delivery is required. This study aimed to develop machine learning models to predict emergency delivery within 48 h of PE diagnosis, facilitating timely clinical interventions. We analyzed a retrospective cohort of 648 singleton pregnancies diagnosed with PE at Fujian Maternal and Child Health Hospital from 2014 to 2024, with gestational ages ranging from 28 to 34 weeks. Patients were stratified into emergency delivery (≤ 48 h post-diagnosis, n = 174) and non-emergency groups (n = 474). Feature selection was performed via univariate analysis, collinearity testing, and logistic regression, yielding 16 predictors. Eight machine learning models (logistic regression, naïve Bayes, XGBoost, LightGBM, SVM, GBDT, MLP, elastic net) were trained and evaluated for discriminative power and calibration. SHAP analysis was employed to interpret model predictions. XGBoost demonstrated superior performance (testing AUROC: 0.908) with excellent calibration, while GBDT achieved high AUROC (0.931) but poorer calibration. SHAP analysis identified CRP, D-dimer, and hypoproteinemia as the most influential predictors. Machine learning models effectively predict emergency delivery in early-onset PE using clinically interpretable features. Integration into obstetric practice may enhance risk stratification, though prospective validation is warranted.