Plasma Olink Proteomics Identifies CCL20 as a Novel Predictive and Diagnostic Inflammatory Marker for Preeclampsia

Using Olink Proteomics to Identify Inflammatory Biomarkers in the Cerebrospinal Fluid in Guillain-Barré Syndrome

Purpose

The precise etiology of Guillain-Barré syndrome (GBS) is uncertain; however, it is linked to immunological and inflammatory processes. Thus, this research aims to investigate new inflammatory biomarkers for GBS diagnosis.

Patients and Methods

In this work, Olink proteomics was used to compare the expression levels of 92 inflammation-related proteins in the cerebrospinal fluid (CSF) of patients with non-inflammatory neurological diseases (n=14) and GBS (n=23). Differentially expressed proteins (DEPs) were then analyzed biologically and in terms of their relationship to clinical features, and logistic regression models were built. We also downloaded GEO data to validate DEPs at the mRNA level.

Results

We identified twenty DEPs. The PPI network screened six key DEPs (including TNF, CCL20, IL8, MCP-1, IL10, and IL5). These DEPs were enriched in the chemokine signaling pathway, the IL-17 signaling pathway, cytokines and their receptor interactions, and other pathways. TNFRSF9 and IL-10RB showed the strongest correlation of expression in CSF. CCL20 and IL5 could be used as potential independent predictors for the diagnosis of GBS. Seven DEPs (MCP-1, CXCL1, MCP-4, MMP-10, CXCL10, CCL28, and CCL20) had some predictive value for the severity of GBS. Based on the validation of the GEO data, the mRNA expression of MCP-1 and CXCL9 was found to be upregulated at the peak of EAN, and the enriched pathways at the gene transcription level were consistent with the results of this study.

Conclusion

DEPs linked to inflammation (such as TNF, CCL20, IL8, MCP-1, IL10, and IL5) could be useful biomarkers for GBS diagnosis. More research is required to determine their precise mechanisms in GBS.

Proteome-Based Maternal Plasma and Serum Biomarkers for Preeclampsia: A Systematic Review and Meta-Analysis

Proteomics has emerged as a transformative tool in biomedical research, enabling comprehensive characterization of protein profiles in complex biological systems. In preeclampsia (PE) research, quantitative proteomic analyses of plasma and serum have revealed critical insights into disease mechanisms and potential biomarkers. Through a systematic review of 17 studies (2009-2024), we identified 561 differentially expressed plasma/serum proteins (p < 0.05) in PE patients versus healthy controls, with 122 proteins consistently replicated across ≥2 independent studies. Stratified analysis by clinical subtype (early-vs. late-onset PE) demonstrated both concordant and divergent protein expression patterns, reflecting heterogeneity in PE pathophysiology, methodological variations (e.g., sample processing, proteomic platforms), and differences between discovery-phase and targeted validation studies. The trimester-specific biomarker panels proposed here offer a framework for future large-scale, multicenter validation. By integrating advanced proteomic technologies with standardized preanalytical and analytical protocols, these findings advance opportunities for early prediction (first-trimester biomarker signatures); mechanistic insight (complement system involvement); and personalized management (subtype-specific therapeutic targets). This work underscores the potential of proteomics to reshape PE research, from molecular discovery to clinical translation, ultimately improving outcomes for this leading cause of maternal and perinatal morbidity.

Seasonal Proteomic Variations and Biomarkers in Seasonal Allergic Rhinitis: Insights from Olink Inflammation Profiling

Purpose

Seasonal allergic rhinitis (SAR) is a prevalent inflammatory condition, yet its molecular mechanisms and reliable biomarkers remain incompletely understood. This study aimed to identify key inflammation-related proteins and pathways associated with SAR by investigating seasonal proteomic profile variations and their correlations with SAR symptoms.

Patients and Methods

Serum samples were collected from nineteen SAR patients during both allergy (in-season, IS) and non-allergy (out-of-season, OS) periods. Differentially expressed proteins (DEPs) were identified using the Olink Target 96 Inflammation panel, which were further analyzed through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. Spearman correlation analysis was conducted to explore associations between DEPs and SAR symptoms, including sneezing, rhinorrhea, nasal blockage, itchy nose, and itchy eye.

Results

A total of 36 inflammation-related DEPs were identified, all significantly upregulated in the allergy season. Notable proteins such as glial cell line-derived neurotrophic factor (GDNF), interleukin-18 receptor 1 (IL-18R1), and interleukin-15 receptor alpha (IL-15RA) showed strong correlations with SAR symptoms. Sneezing was associated with IL-2 receptor beta (IL-2RB) (r = 0.415, p = 0.013), rhinorrhea with FMS-related tyrosine kinase 3 ligand (Flt3L) (r = 0.455, p = 0.004), and nasal blockage with osteoprotegerin (OPG) (r = 0.493, p = 0.002). GO analysis revealed enrichments in Ras signaling and small GTPase pathways, while KEGG analysis highlighted immune-related pathways, including PI3K-Akt signaling and cytokine-cytokine receptor interactions.

Conclusion

This study identified key inflammation-related proteins and pathways that vary seasonally in SAR, offering insights into potential biomarkers and therapeutic targets for SAR management. Further studies are recommended to validate these findings in larger and more diverse populations.

Targeted plasma proteomics reveals organ damage signatures of AIDS- and noncommunicable disease-related deaths in people with HIV

Antiretroviral therapy (ART) is shifting the primary driver of mortality for people with HIV (PWH) from opportunistic infections to noncommunicable diseases (NCDs). Protein biomarkers differentiating both AIDS-related and NCDs-related deaths from PWH may help early and precise risk prediction and intervention. We conduct a nested case-control study where 126 HIV deaths, 162 age-sex-matched HIV survivors and 152 HIV-negative controls are analyzed with 92 protein biomarkers of the Olink Organ Damage panel by proximity extension assays (PEA). Using LASSO regression, logistic regression, and ROC analysis, twelve proteins are significantly associated with HIV death, of which six (SIRT5, PPM1B, PSMA1, GALNT10, VEGFC, PTN) are specifically associated with NCDs-related death, two (RCOR1, SERPINA9) are specifically associated with AIDS-related death, and four (CA12, CA14, RARRES1, EDIL3) are associated with both. Three of these proteins are replicable in the external validation sample. The adjusted protein panels consisting of significantly associated proteins selected through both LASSO and logistic regression model well predicted NCDs-related death (AUC = 0.970) and AIDS-related death (AUC = 0.960) in PWH. The selected proteins also displayed a significant correlation with traditional biomarkers of NCDs among PWH (P < 0.05). The potential clinical utility of these biomarkers could shed light on pathogenesis of end-stage organ dysfunction in PWH.

Large-scale plasma proteomics uncovers novel targets linking ambient air pollution and depression

Despite the growing recognition of association between air pollution and increased risk of depression, the intricate biological mechanisms underlying it remains unclear. In this study, a total of 1463 plasma proteins were measured by the Olink Explore platform for 50,553 participants in a large prospective cohort. Four air pollutants were assessed using land-use regression models: particulate matter with aerodynamic diameter ≤ 2.5μm (PM2.5), particulate matter with aerodynamic diameter > 2.5μm and ≤ 10μm (PM2.5-10), nitrogen dioxide (NO2), and nitric oxide (NO). The air pollution index was calculated using principal components analysis to assess joint exposure to air pollution. Logistic regression and Cox proportional hazards regression analyses were respectively used to explore the impact of the interaction between air pollution exposure and plasma proteins on the prevalence and incidence of depression. Functional enrichment analysis and drug prediction analysis were conducted to explore the biological mechanisms and drugs associated with identified plasma proteins with interaction effects. Logistic regression analysis detected seven significant air pollutant and plasma protein interactions for the prevalence of depression, such as CDHR5 vs. PM2.5 (OR: 0.58; 95% CI: 0.48-0.71), TNFRSF13C vs. NO (OR :0.70, 95% CI: 0.58-0.84) and ICAM5 vs. air pollution index (OR: 1.38, 95% CI: 1.17-1.63). Two significant interactions were identified for the incidence of depression: CDHR5 vs. PM2.5 (HR: 0.62, 95% CI: 0.50-0.76) and HSD11B1 vs. PM2.5 (HR: 1.48, 95% CI: 1.22-1.81). The plasma proteins that interacted with air pollutants were enriched in various Gene Ontology terms and pathways involving immunity, endocrine, inflammation, neurological function and metabolism, such as neuroinflammatory response, neuron projection guidance, regulation of lymphocyte mediated immunity, steroid biosynthetic process and lipid digestion. We also found that these proteins interacted with multiple drugs, such as risperidone, olanzapine and progesterone. This study identified novel targets linking ambient air pollution and depression, providing the insights for biological mechanisms of air pollution affecting the risk of depression.

Olink Profiling of Intestinal Tissue Identifies Novel Biomarkers For Colorectal Cancer

Comprehensive protein profiling in intestinal tissues provides detailed information about the pathogenesis of colorectal cancer (CRC). This study quantified the expression levels of 92 oncology-related proteins in tumors, paired para-carcinoma tissues, and remote normal tissues from a cohort of 52 CRC patients utilizing the Olink technology. The proteomic profile of normal tissues closely resembled that of para-carcinoma tissues while distinctly differing from that of tumors. Among the 68 differentially expressed proteins (DEPs) identified between the tumor and normal tissues, WISP-1, ESM-1, and TFPI-2 showed the most pronounced alterations and exhibited relatively strong correlations. These markers also presented the highest AUC values for distinguishing between tissue types. Bioinformatic analysis of the DEPs revealed that the plasma membrane and the PI3K-AKT signaling pathway were among the most enriched GO terms and KEGG pathways. Furthermore, although TFPI-2 is typically recognized as a tumor suppressor, both Olink and enzyme linked immunosorbent assay (ELISA) analyses have demonstrated that its expression is significantly elevated in tumors compared with paired normal tissues. To the best of our knowledge, this is the first study to profile the proteome of intestinal tissue using the Olink technology. This work offers valuable insights into potential biomarkers and therapeutic targets for CRC, complementing the Olink profiling of circulating proteins.

Combined Plasma Olink Proteomics and Transcriptomics Identifies CXCL1 and TNFRSF12A as Potential Predictive and Diagnostic Inflammatory Markers for Acute Kidney Injury

Acute kidney injury (AKI) poses a significant global public health challenge. Current methods for detecting AKI rely on monitoring changes in serum creatinine (Scr), blood urea nitrogen (BUN), urinary output and some commonly employed biomarkers. However, these indicators are usually neither specific nor sensitive to AKI, especially in cases of mild kidney injury. AKI is accompanied by severe inflammatory reactions, resulting in the upregulation of numerous inflammation-associated proteins in the plasma. Plasma biomarkers are a noninvasive method for detecting kidney injury, and to date, plasma inflammation-associated cytokines have not been adequately studied in AKI patients. The objective of our research was to identify novel inflammatory biomarkers for AKI. We utilized Olink proteomics to analyze the alterations in plasma inflammation-related proteins in the serum of healthy mice (n = 2) or mice treated with cisplatin (n = 6). Additionally, transcriptome datasets for the lipopolysaccharide (LPS), cisplatin, and ischemia‒reperfusion injury (IRI) groups were obtained from the National Center of Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database. We calculated the intersection of differentially expressed proteins (DEPs) and genes (DEGs) from both datasets. In the Olink proteomics analysis, the AKI group had significantly greater levels of 11 DEPs than did the control group. In addition, 56 common upregulated DEGs were obtained from the transcriptome dataset. The expression of CXCL1 and TNFRSF12A overlapped across all the datasets. The transcription and protein expression levels of CXCL1 and TNFRSF12A were detected in vivo. The gene and protein levels of CXCL1 and TNFRSF12A were significantly increased in different AKI mouse models and clinical patients, suggesting that these genes and proteins could be potential specific biomarkers for the identification of AKI.

Plasma Proteomics Analysis of Early Biomarkers for Predicting Female Fecundability: A Nested Case-Control Study

The present study aimed to identify and verify new plasma protein markers to predict the female fecundability level. A nested case-control study was conducted involving couples who participated in the Chinese National Free Preconception Check-up Project. Women who successfully conceive within one year were defined as the high fecundability group, and those unable to conceive were defined as the low fecundability group. In the training cohort, potential protein biomarkers were identified using proteomics technology and were further tested in a validation cohort by the Western blotting assay, enzyme-linked immunosorbent assay, and biochemical tests. Meanwhile, receiver operating characteristic curve analysis were used to evaluate the predictive value. Cox proportional hazard regression analyses were conducted to calculate hazard ratios; restricted cubic spline analysis was used to assess the linear relationship between the the protein level and hazard ratios for fecundability. Pyruvate, a key product of glycolysis, was significantly increased in the high fecundability group (P < 0.01) compared to the low fecundability group, and its area under the curve value was 0.68 (P < 0.05). There was a linear positive dose-response association between the pyruvate level and fecundability possibility (hazard ratios = 1.66, 95% CI: 1.07-2.59, p for trend = 0.025, nonlinearity, p-value = 0.2927).

Methods and clinical biomarker discovery for targeted proteomics using Olink technology

PURPOSE

This paper is to offer insights for designing research utilizing Olink technology to identify biomarkers and potential therapeutic targets for disease treatment.

EXPERIMENTAL DESIGN

We discusses the application of Olink technology in oncology, cardiovascular, respiratory and immune-related diseases, and Outlines the advantages and limitations of Olink technology.

RESULTS

Olink technology simplifies the search for therapeutic targets, advances proteomics research, reveals the pathogenesis of diseases, and ultimately helps patients develop precision treatments.

CONCLUSIONS

Although proteomics technology has been rapidly developed in recent years, each method has its own disadvantages, so in the future research, more methods should be selected for combined application to verify each other.

Integration of proteomics with prospective birth cohort to elucidate early life origins of cardiometabolic diseases: rationale, study design, lab assay, and quality control

There is growing evidence that the plasma proteome provides insights into personal health status at different stages of life. However, limited data are available on high-throughput proteomic studies in pediatric populations, especially, using prospective birth cohorts. We launched a proteomics study in 990 children from a US predominantly urban, low-income, multi-ethnic prospective Boston Birth Cohort (BBC, referred as "BBC proteomics study"), which aimed to leverage proteomics to investigate the biological pathways underlying the link between preterm birth and child long-term cardiometabolic health. The objective of this paper is to describe the rationale, study design, proteomic assay and quality control steps for the BBC proteomics study in a subset of children with available proteomic profiling. Using the OLINK® Explore 3072 platform, proteomic profiling was performed in cord plasma at birth and in postnatal plasma collected during early childhood. Quality control (QC) steps were performed, including calculation of coefficient of variation (CV), missingness rates per sample or per protein, principal component analyses to identify clustering and outliers, and correlation analyses among the duplicates to indicate reproducibility. A total of 2,941 proteins from eight OLINK panels were successfully measured at both time points. Almost 100% of samples passed lab-prespecified QC. Approximately 89% of proteins were detected in > 50% samples; 79.6% had intra-CV < 15% and 79.9% of had inter-CV < 30%. Four samples were identified as outliers due to high missingness rates. Our data also demonstrated that this assay had a good reproducibility with correlation coefficient (r) > 0.65 in most of the duplicates, although we also identified potential batch effects. In conclusion, our data suggests that this high-throughput proteomic profiling is feasible and reproducible in archived plasma samples, including cord blood. We anticipated that successful completion of this proteomics study will help identify novel predictive biomarkers and therapeutic targets so that high-risk newborns can be identified, and effective interventions can be initiated during the earliest developmental window when they may have the greatest life-long benefit.

MUC16 can Predict the Pregnancy Outcomes in Human and Intraperitoneal Administration of MUC16 can Rescue Pregnancy Losses in Mouse Models

Mucin 16 (MUC16) participates in the process of embryo implantation, but few studies have examined the association between MUC16 and pregnancy loss. To investigate this association, the expression of MUC16 in serum and decidua was compared between women with pregnancy loss and ongoing pregnancies. In vitro experiments and animal models were used to explore the role and underlying mechanisms of MUC16 in pregnancy loss. In human study, the expression of MUC16 in serum and decidua was both consistently lower in the women with pregnancy loss compared with those in women with ongoing pregnancies. In vitro experiments revealed the interaction of MUC16 with peripheral blood natural killer (pNK) cells. MUC16 changed the phenotype and reduced the pro-inflammation ability of pNK cells. MUC16 also inhibited the cytotoxicity of pNK cells through the Src homology region 2 domain-containing phosphatase-1/extracellular signal-regulated kinase (SHP-ERK) pathway. Furthermore, MUC16 promoted the migration, invasion and tube formation of trophoblast cells by co-culturing together with pNK cells. In vivo experiments, the mouse model of abortion was used to further confirm that intraperitoneal administration of MUC16 could rescue the pregnancy loss. This study reveals the still-unknown connection between MUC16 and pNK cells and indicates that MUC16 provides a novel method for future prediction and treatment of unfavorable pregnancy outcomes.

Behind the Curtain of Abnormal Placentation in Pre-Eclampsia: From Molecular Mechanisms to Histological Hallmarks

Successful human pregnancy needs several highly controlled steps to guarantee an oocyte's fertilization, the embryo's pre-implantation development, and its subsequent implantation into the uterine wall. The subsequent placenta development ensures adequate fetal nutrition and oxygenation, with the trophoblast being the first cell lineage to differentiate during this process. The placenta sustains the growth of the fetus by providing it with oxygen and nutrients and removing waste products. It is not surprising that issues with the early development of the placenta can lead to common pregnancy disorders, such as recurrent miscarriage, fetal growth restriction, pre-eclampsia, and stillbirth. Understanding the normal development of the human placenta is essential for recognizing and contextualizing any pathological aberrations that may occur. The effects of these issues may not become apparent until later in pregnancy, during the mid or advanced stages. This review discusses the process of the embryo implantation phase, the molecular mechanisms involved, and the abnormalities in those mechanisms that are thought to contribute to the development of pre-eclampsia. The review also covers the histological hallmarks of pre-eclampsia as found during the examination of placental tissue from pre-eclampsia patients.

The Olink proteomics profile in nasal secretion of patients with allergic rhinitis

KEY POINTS

Nasal secretions of allergic rhinitis patients were analyzed by Olink proteomics. Fifteen differentially expressed proteins (DEPs) were identified. The DEPs were significantly correlated with the total nasal symptom scores of patients with allergic rhinitis.

Genetic Associations of Circulating Cardiovascular Proteins With Gestational Hypertension and Preeclampsia

Importance

Hypertensive disorders of pregnancy (HDPs), including gestational hypertension and preeclampsia, are important contributors to maternal morbidity and mortality worldwide. In addition, women with HDPs face an elevated long-term risk of cardiovascular disease.

Objective

To identify proteins in the circulation associated with HDPs.

Design, Setting, and Participants

Two-sample mendelian randomization (MR) tested the associations of genetic instruments for cardiovascular disease-related proteins with gestational hypertension and preeclampsia. In downstream analyses, a systematic review of observational data was conducted to evaluate the identified proteins' dynamics across gestation in hypertensive vs normotensive pregnancies, and phenome-wide MR analyses were performed to identify potential non-HDP-related effects associated with the prioritized proteins. Genetic association data for cardiovascular disease-related proteins were obtained from the Systematic and Combined Analysis of Olink Proteins (SCALLOP) consortium. Genetic association data for the HDPs were obtained from recent European-ancestry genome-wide association study meta-analyses for gestational hypertension and preeclampsia. Study data were analyzed October 2022 to October 2023.

Exposures

Genetic instruments for 90 candidate proteins implicated in cardiovascular diseases, constructed using cis-protein quantitative trait loci (cis-pQTLs).

Main Outcomes and Measures

Gestational hypertension and preeclampsia.

Results

Genetic association data for cardiovascular disease-related proteins were obtained from 21 758 participants from the SCALLOP consortium. Genetic association data for the HDPs were obtained from 393 238 female individuals (8636 cases and 384 602 controls) for gestational hypertension and 606 903 female individuals (16 032 cases and 590 871 controls) for preeclampsia. Seventy-five of 90 proteins (83.3%) had at least 1 valid cis-pQTL. Of those, 10 proteins (13.3%) were significantly associated with HDPs. Four were robust to sensitivity analyses for gestational hypertension (cluster of differentiation 40, eosinophil cationic protein [ECP], galectin 3, N-terminal pro-brain natriuretic peptide [NT-proBNP]), and 2 were robust for preeclampsia (cystatin B, heat shock protein 27 [HSP27]). Consistent with the MR findings, observational data revealed that lower NT-proBNP (0.76- to 0.88-fold difference vs no HDPs) and higher HSP27 (2.40-fold difference vs no HDPs) levels during the first trimester of pregnancy were associated with increased risk of HDPs, as were higher levels of ECP (1.60-fold difference vs no HDPs). Phenome-wide MR analyses identified 37 unique non-HDP-related protein-disease associations, suggesting potential on-target effects associated with interventions lowering HDP risk through the identified proteins.

Conclusions and Relevance

Study findings suggest genetic associations of 4 cardiovascular disease-related proteins with gestational hypertension and 2 associated with preeclampsia. Future studies are required to test the efficacy of targeting the corresponding pathways to reduce HDP risk.

Vitreous Olink proteomics reveals inflammatory biomarkers for diagnosis and prognosis of traumatic proliferative vitreoretinopathy

Background

The aim of this study was to identify inflammatory biomarkers in traumatic proliferative vitreoretinopathy (TPVR) patients and further validate the expression curve of particular biomarkers in the rabbit TPVR model.

Methods

The Olink Inflammation Panel was used to compare the differentially expressed proteins (DEPs) in the vitreous of TPVR patients 7-14 days after open globe injury (OGI) (N = 19) and macular hole patients (N = 22), followed by correlation analysis between DEPs and clinical signs, protein-protein interaction (PPI) analysis, area under the receiver operating characteristic curve (AUC) analysis, and function enrichment analysis. A TPVR rabbit model was established and expression levels of candidate interleukin family members (IL-6, IL-7, and IL-33) were measured by enzyme-linked immunosorbent assay (ELISA) at 0, 1, 3, 7, 10, 14, and 28 days after OGI.

Results

Forty-eight DEPs were detected between the two groups. Correlation analysis showed that CXCL5, EN-RAGE, IL-7, ADA, CD5, CCL25, CASP8, TWEAK, and IL-33 were significantly correlated with clinical signs including ocular wound characteristics, PVR scoring, PVR recurrence, and final visual acuity (R = 0.467-0.699, p < 0.05), and all with optimal AUC values (0.7344-1). Correlations between DEP analysis and PPI analysis further verified that IL-6, IL-7, IL-8, IL-33, HGF, and CXCL5 were highly interactive (combined score: 0.669-0.983). These DEPs were enriched in novel pathways such as cancer signaling pathway (N = 14, p < 0.000). Vitreous levels of IL-6, IL-7, and IL-33 in the rabbit TPVR model displayed consistency with the trend in Olink data, all exhibiting marked differential expression 1 day following the OGI.

Conclusion

IL-7, IL-33, EN-RAGE, TWEAK, CXCL5, and CD5 may be potential biomarkers for TPVR pathogenesis and prognosis, and early post-injury may be an ideal time for TPVR intervention targeting interleukin family biomarkers.

Can single-cell and spatial omics unravel the pathophysiology of pre-eclampsia?

Pre-eclampsia is a leading cause of maternal and fetal morbidity and mortality. Characterised by the onset of hypertension and proteinuria in the second half of pregnancy, it can lead to maternal end-organ injury such as cerebral ischemia and oedema, pulmonary oedema and renal failure, and potentially fatal outcomes for both mother and fetus. The causes of the different maternal end-organ phenotypes of pre-eclampsia and why some women develop pre-eclampsia condition early in pregnancy have yet to be elucidated. Omics methods include proteomics, genomics, metabolomics, transcriptomics. These omics techniques, previously mostly used on bulk tissue and individually, are increasingly available at a single cellular level and can be combined with each other. Multi-omics techniques on a single-cell or spatial level provide us with a powerful tool to understand the pathophysiology of pre-eclampsia. This review will explore the status of omics methods and how they can and could contribute to understanding the pathophysiology of pre-eclampsia.

A cost-effective machine learning-based method for preeclampsia risk assessment and driver genes discovery

BACKGROUND

The placenta, as a unique exchange organ between mother and fetus, is essential for successful human pregnancy and fetal health. Preeclampsia (PE) caused by placental dysfunction contributes to both maternal and infant morbidity and mortality. Accurate identification of PE patients plays a vital role in the formulation of treatment plans. However, the traditional clinical methods of PE have a high misdiagnosis rate.

RESULTS

Here, we first designed a computational biology method that used single-cell transcriptome (scRNA-seq) of healthy pregnancy (38 wk) and early-onset PE (28-32 wk) to identify pathological cell subpopulations and predict PE risk. Based on machine learning methods and feature selection techniques, we observed that the Tuning ReliefF (TURF) score hybrid with XGBoost (TURF_XGB) achieved optimal performance, with 92.61% accuracy and 92.46% recall for classifying nine cell subpopulations of healthy placentas. Biological landscapes of placenta heterogeneity could be mapped by the 110 marker genes screened by TURF_XGB, which revealed the superiority of the TURF feature mining. Moreover, we processed the PE dataset with LASSO to obtain 497 biomarkers. Integration analysis of the above two gene sets revealed that dendritic cells were closely associated with early-onset PE, and C1QB and C1QC might drive preeclampsia by mediating inflammation. In addition, an ensemble model-based risk stratification card was developed to classify preeclampsia patients, and its area under the receiver operating characteristic curve (AUC) could reach 0.99. For broader accessibility, we designed an accessible online web server ( http://bioinfor.imu.edu.cn/placenta ).

CONCLUSION

Single-cell transcriptome-based preeclampsia risk assessment using an ensemble machine learning framework is a valuable asset for clinical decision-making. C1QB and C1QC may be involved in the development and progression of early-onset PE by affecting the complement and coagulation cascades pathway that mediate inflammation, which has important implications for better understanding the pathogenesis of PE.