Advancing Sepsis Diagnosis and Immunotherapy Machine Learning-Driven Identification of Stable Molecular Biomarkers and Therapeutic Targets

Advancing Sepsis Diagnosis and Immunotherapy Machine Learning-Driven Identification of Stable Molecular Biomarkers and Therapeutic Targets | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Advancing Sepsis Diagnosis and Immunotherapy Machine Learning-Driven Identification of Stable Molecular Biomarkers and Therapeutic Targets Fangpeng Liu, Weichuan Xiong, Rui Xiao, Yian Zhan This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4306022/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This research presents a novel integrated approach combining genomic analysis and machine learning to identify biomarkers and drug sensitivities specific to sepsis, aiming to facilitate personalized treatment strategies. We comprehensively examined gene expression profiles from sepsis patients and healthy controls by utilizing the Gene Expression Omnibus (GEO) database, specifically datasets GSE154918 and GSE134347. Through the application of the ESTIMATE algorithm, weighted gene co-expression network analysis (WGCNA), and a range of machine learning techniques, we identified crucial Sepsis-Related Genes (SRGs), Immune-Related Differentially Expressed Genes (IRDEGs), and Important Immune-related genes (IIRGs). Our analysis revealed significant differences in immune and stromal scores between sepsis patients and controls, highlighting the altered immune landscape in sepsis. The study also uncovered specific genes associated with drug sensitivity, providing insights into potential therapeutic targets. The predictive model developed demonstrated high accuracy in sepsis diagnosis and prognosis, validated by independent datasets. These findings contribute to understanding sepsis at a molecular level and open new avenues for developing personalized therapeutic interventions, underscoring the potential of integrating genomic analysis and machine learning in sepsis research. Biological sciences/Biotechnology/Genomics/Phylogenomics Biological sciences/Drug discovery/Biomarkers/Diagnostic markers Biological sciences/Molecular biology/Proteomics/Protein–protein interaction networks Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Introduction Sepsis, a critical condition stemming from the body's maladaptive response to infection, is characterized by its high prevalence and mortality rate, exerting a profound impact on healthcare systems worldwide. A landmark study featured in The Lancet in August 2020, titled 'Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study' 1 , provided a comprehensive evaluation of sepsis's global impact. This investigation highlighted an alarming figure of approximately 48.9 million sepsis cases and 11 million resultant fatalities worldwide in 2017, numbers that significantly surpass previous assessments. The findings of this study brought to light the criticality of sepsis as a leading cause of death globally, presenting it as a major public health issue. It also stressed the urgency of rapid and accurate diagnosis, coupled with effective treatment strategies, as pivotal factors in improving patient survival rates. Sepsis manifests through a disordered immune reaction, encompassing both the innate and adaptive immune systems. This disorder often leads to an overstimulation of immune cells, causing rampant inflammation and subsequent tissue harm. Current investigations in the field of sepsis are concentrating on the identification of biomarkers, including widely recognized ones 2 , 3 , 4 , like C-reactive protein (CRP), Procalcitonin (PCT), Interleukin 6 (IL-6), pro-inflammatory cytokines (e.g., TNF-α, IL-1β), soluble Triggering Receptor Expressed on Myeloid cells-1 (sTREM-1) , endothelial biomarkers, MicroRNA, and neutrophil CD64 . These biomarkers, emerging from a combination of fundamental and clinical research coupled with technological progress, have yet to provide a definitive diagnostic or prognostic capability for sepsis. Although helpful in diagnosis and prognostication, many of these markers still fall short in terms of distinctiveness or clinical applicability in the treatment of sepsis 4 , 5 ." As medical science evolves swiftly, machine learning models emerge as groundbreaking approaches, potentially outperforming conventional methods in the early and ongoing detection of sepsis. This advancement could lead to timelier interventions, thereby improving patient treatment outcomes 6 . These models, founded on data-driven algorithms, reduce reliance on the subjective aspects of sepsis diagnosis, which traditionally hinge on clinical discretion 7 . Machine learning has established itself as a crucial component in medical diagnostics. It excels in discerning intricate patterns in biological data, patterns that frequently pose challenges to human experts. This is achieved through the utilization of expansive datasets and sophisticated algorithms 8 . Contemporary studies in machine learning for sepsis prediction are making notable strides. One notable research 9 has put forth the SepsisFinder algorithm, showcasing its proficiency in predicting sepsis earlier compared to other models like NEWS2 and GBDT under equivalent sensitivity configurations. This finding elevates SepsisFinder as a frontrunner in early sepsis detection. Additionally, a comprehensive systematic review and meta-analysis 10 explored the capacity of machine learning to forecast sepsis-induced mortality. This analysis, incorporating diverse machine learning models, has confirmed their formidable potential in this particular domain. A focused meta-analysis 6 on the use of machine learning for sepsis onset prediction in ICU settings revealed the Random Forest (RF) model's superior accuracy and the eXtreme Gradient Boosting (XGBoost) model's exemplary predictive performance. These insights affirm the significant role of machine learning in medical practice, especially for early detection and prediction of sepsis. Our study intends to amalgamate various algorithms with extensive patient data, employing multifaceted classifier machine-learning models to refine prediction precision and reliability. We are committed to augmenting sepsis early detection and enhancing intervention efficiency at crucial immune monitoring junctures. This article delves into the utility of machine learning in deriving biomarkers from immune sources for sepsis and clarifying the potential roles of immune cells in sepsis etiology. We introduce a unique multi-classifier machine learning model for sepsis, breaking away from traditional logistic regression or LASSO regression (Least Absolute Shrinkage and Selection Operator Regression) modeling approaches. Our focus is on immune-related genes, constructing diverse predictive models to boost sepsis diagnostic accuracy. By evaluating the significance of each gene through feature importance scores or SHAP values, we analyze their mechanisms related to immune cells and their correlation with drug sensitivity. This approach is pivotal in disease diagnosis, monitoring progression, evaluating treatment efficacy, and informing clinical decisions. It holds the promise of facilitating early diagnosis, customizing treatment plans, propelling personalized medicine, and ultimately elevating patient care in various medical scenarios. Materials and methods 1.1 Data Download Utilizing the R package GEOquery (version 2.68.0) 11 ,we accessed the GEO database to retrieve expression data from the GSE154918 (n = 105) and GSE134347 (n = 298) 12 datasets, both of which originated from Homo sapiens. The GSE154918 dataset comprises gene expression profiles from 56 patients clinically diagnosed with Sepsis and 49 healthy control subjects, totaling 105 samples, which were used as validation dataset in this project. This dataset provides a comprehensive comparison between the gene expression patterns in patients with Sepsis and those in healthy individuals. Similarly, the GSE134347 dataset includes gene expression data from 215 Sepsis patients and 83 healthy controls, amounting to 298 samples in total, which were used as the training dataset in this project.This dataset offers an extensive resource for analyzing the genetic underpinnings of Sepsis, contrasting the expression profiles of affected individuals with those of healthy donors. The data platform of dataset GSE154918 was GPL20301 Illumina HiSeq 4000 (Homo sapiens); the data platform of dataset GSE134347 is GPL17586 Affymetrix Human Transcriptome Array 2. All sample data in dataset GSE154918 and GSE134347 were removed from batch effect for follow-up analysis. See Table 1 for specific dataset information. Table 1 List of datasets used in this study. Datasets Name Source No. Of Sample GSE154918 GEO 298 GSE134347 GEO 105 To gather Immune-Related Genes (IRGs), we utilized the ImmPort database, accessible at [ https://www.immport.org/home ]. ImmPort serves as a pivotal resource in the field of immunology, offering a platform for the aggregation, organization, and dissemination of research data. It supports the life sciences research community by enabling the archival and exchange of scientific data through advanced information technology. This database not only provides a robust repository for research data but also ensures the long-term, sustainable storage of both research and clinical data. From this database, we meticulously compiled and cross-verified a list of 1509 unique IRGs. For detailed information on these genes, refer to Supplementary Table S1 in our manuscript. 1.2 Estimation Estimation (Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data) is an algorithm used to ESTIMATE the purity of a sample tumor. Developed and maintained by MD Anderson Cancer. A simple gene expression matrix is input to infer the content levels of immune cells, stromal cells, and tumor purity of the sample. Therefore, we input the sepsis dataset GSE134347 into the ESTIMATE algorithm to calculate the immune score, stroma score, ESTIMATE score and tumor purity of each sample, and then test the differences of these four scores in different sepsis and non-sepsis subgroups by Wilcoxon. P < 0.05 was considered statistically significant. 1.3 WGCNA algorithm was used to identify disease-related genes in the sepsis dataset. To isolate Sepsis-Related Genes (SRGs), our initial step involved the application of Weighted Gene Correlation Network Analysis (WGCNA) 13 . The purpose of WGCNA is to discern modules of co-expressed genes, elucidate the nexus between gene networks and immunity, and identify pivotal genes within these networks. We utilized the 'pickSoftThreshold' function to determine the optimal soft threshold, which was found to be 5, facilitating the construction of scale-free networks based on this threshold. Subsequently, topological matrices were generated, followed by hierarchical clustering. Setting 50 as the minimum gene count for each module, we dynamically sliced and identified gene modules, and computed module Eigengenes. Module Eigengenes were then used to establish inter-module correlations and to perform further hierarchical clustering. Modules exhibiting correlations above 0.25 were amalgamated, resulting in a total of 22 distinct modules. The relationship between these modules and clinical features was analyzed using Pearson or Spearman correlation analysis. To acquire immune-related differentially expressed genes (IRDEGs) linked to sepsis, we intersected the SRGs identified through WGCNA with the IRGs sourced from the ImmPort database. To visualize these genes, we employed the R package 'pheatmap' (version 1.0.12) to create an expression heatmap. 1.4 Screening for important immune-related genes in the sepsis dataset To refine the identification of important genes within IRDEGs, we employed five prevalent Machine Learning algorithms, namely Elastic Net, LASSO regression, RF, Boruta, and XGBoost decision trees. The Elastic Net is a linear regression model that incorporates both L1 and L2 norm regularization in its training dataset. LASSO regression, based on linear regression, introduces a penalty term (lambda × absolute value of coefficient) to reduce model overfitting and enhance generalizability. Both Elastic Net and LASSO regression were implemented using the R package 'glmnet'. The outcomes of LASSO regression are visually represented through diagnostic model diagrams and variable trace plots. RF employs ensemble learning to integrate multiple decision trees. As part of the bagging (bootstrap aggregation) algorithm within ensemble learning, it is executed via the 'caret' package. RF functions by aggregating predictions from multiple trees, with the final decision derived through a majority vote. Boruta, a feature selection method gaining popularity, identifies all features correlated with the dependent variable, regardless of their impact on a specific model's cost function. We applied the 'Boruta' package to achieve this. XGBoost, a gradient boosting algorithm, builds its model iteratively, each time adding a CART tree that fits the residual differences from the previous trees’ predictions. This is facilitated by the 'xgboost' package. Additionally, each machine learning algorithm is fine-tuned using Cross-Validation (CV) for hyperparameter optimization, ensuring model performance enhancement. To bolster robustness, we repeated the optimization ten times for each resampling, each time with a different random seed. Ultimately, to achieve stable results, genes identified by all five machine learning algorithms were consolidated as the final set of important immune-related genes(IIRGs)for our ensuing predictive model development. 1.5 The diagnostic model of IIRGs was constructed in the sepsis dataset In our pursuit to develop a sophisticated sepsis response classification model, we harnessed the potential of IIRGs by training them with an array of six prevalent machine learning algorithms. This suite included Naive Bayes (NB), Conditional Inference Random Forest(cforest), LogitBoost (an advanced form of logistic regression), Gradient Boosting Machine (GBM), Model Averaged Neural Network (avNNet), and Penalized Discriminant Analysis (pda). For all these machine learning algorithms, we meticulously employed CV as a method for hyperparameter tuning, aiming to enhance the model's performance and accuracy. To assure the robustness of our models, we diligently repeated this optimization process ten times, each with a unique random seed for each iteration of resampling. Following the construction of classifiers using these diverse algorithmic models, we conducted a thorough analysis through validation dataset GSE154918. This step was crucial in determining the most effective algorithm in terms of classification performance within the validation dataset. Subsequently, the algorithm demonstrating the best classification efficacy was chosen for the final assembly of our sepsis prediction model. This approach underscores our commitment to precision and reliability in developing a model adept at predicting sepsis with high accuracy." 1.6 Enrichment analysis The utilization of GO analysis stands as a prevalent approach for conducting comprehensive functional enrichment studies. This analysis encompasses three key areas: Molecular Function (MF), Biological Process (BP) and Cellular Component (CC) 14 . Additionally, the KEGG database is extensively employed for its vast repository of information on genomes, biological pathways, diseases, and pharmaceuticals 15 . To delve into the potential mechanisms underlying the action of the identified crucial immune-related genes, we leveraged the 'clusterProfiler' package in R (version 4.8.3). This package facilitated our detailed exploration through GO annotation analysis and KEGG pathway enrichment analysis. In our study, a False Discovery Rate (FDR) threshold of less than 0.05 was set as the benchmark for statistical significance, ensuring the reliability and relevance of our findings. 1.7 CIBERSORT CIBERSORT (available at [ https://cibersortx.stanford.edu/ ]) operates on the principle of linear support vector regression, a sophisticated statistical method used in machine learning 16 . This tool, available both as an R package and a web-based application, specializes in the deconvolution of expression matrices of various human immune cell subtypes. It is particularly adept at evaluating the infiltration status of immune cells in sequenced samples, utilizing a gene expression signature set representative of 22 distinct immune cell subtypes. In our research, we employed the CIBERSORT algorithm to assess the infiltration of immune cells in a composite dataset comprising different tumor samples. We then applied the Wilcoxon test to analyze the variance in immune cell infiltration between different sepsis and non-sepsis subgroups. For our study, a P-value of less than 0.05 was set as the threshold for statistical significance, ensuring the rigor and validity of our findings. 1.8 ssGSEA immunoinfiltration analysis The single-sample gene set enrichment analysis (ssGSEA) algorithm was implemented to accurately quantify the relative abundance of each type of immune cell infiltration 17 . Initially, specific labels were assigned to various infiltrated immune cell types, such as Activated CD8 T cell, Gamma delta T cell, Natural killer cell, and Regulatory T cell, among other human immune cell subtypes. The ssGSEA analysis then computed enrichment scores, which served as indicators of the relative abundance of each immune cell type within individual samples. To visually represent these findings, we utilized the ggplot2 package (version 3.4.2) for graphical illustration of the distribution patterns in both sepsis and control groups. Additionally, the Wilcoxon test was employed to ascertain the differences in immune cell infiltration between sepsis and non-sepsis subgroups. In our analysis, a P-value threshold of less than 0.05 was established to denote statistical significance, thereby ensuring the robustness and validity of our results. 1.9 Correlation analysis of IIRGs and immune infiltration in the sepsis dataset To delve deeper into the underlying mechanisms of IIRGs in sepsis, our study extended its analysis to the correlation between the expression of these IIRGs and immune infiltration in sepsis patients, specifically within the sepsis dataset GSE134347. Moreover, we ventured to investigate the relationship between the expression of these IIRGs and the expression of immune checkpoints in sepsis patients. For this correlation analysis, we employed Pearson correlation analysis as our primary analytical tool. To visually represent these correlations, the 'ggcorrplot' R package (version 0.1.4.1) was utilized, enabling us to create detailed and informative correlation loop diagrams. This comprehensive approach allowed us to gain a clearer understanding of the interactions and potential impact of these IIRGs in the context of sepsis and endometriosis. 1.10 PPI Network Analysis (STRING) The STRING database, renowned for its comprehensive mapping of both established and speculative Protein-Protein Interaction (PPI) 18 , served as a pivotal resource in our study. We utilized this database to construct a PPI network for the crucial genes we identified. The parameters for this construction were meticulously set at coefficients of 0.4, 0.7, and 0.9 to ensure optimal specificity and relevance. The data derived from the STRING database were then exported and intricately visualized using Cytoscape 19 , a powerful tool for complex network analysis and visualization. Furthermore, to gain deeper insights into the central components of this network, we employed the CytoHubba plug-in 20 ,for an in-depth analysis of the Hub genes within the PPI network. This comprehensive approach allowed us to unravel the intricate web of interactions among key proteins and identify pivotal genes in the context of our study. 1.11 Drug sensitivity analysis The intricate genomic alterations in various cancers play a pivotal role in influencing clinical treatment responses, often acting as reliable biomarkers for drug efficacy. In this context, the Genomics of Drug Sensitivity in Cancer (GDSC) database (accessible at www.cancerRxgene.org ) stands as the most extensive publicly available repository, offering valuable insights into drug sensitivity and molecular indicators of drug response in cancer cells. This database is instrumental in uncovering data on tumor drug responses and identifying genomic markers linked to drug sensitivity. In our research, we harnessed the capabilities of the pRRophetic algorithm. This tool enables us to predict the sensitivity of patients, categorized into different clinical variable groups, to prevalent anti-cancer drugs or small molecular compounds. This prediction is based on the analysis of expression matrices from the dataset and involves calculating IC50 values. To effectively present our findings, we utilized comparative group comparison graphs, which allow a clear and detailed visual representation of the results, thereby facilitating a better understanding of the drug response dynamics in cancer treatments. 1.12 Statistical analysis Our study's data processing and statistical analysis were meticulously conducted using the R software (available at [ https://www.r-project.org/ ], version 4.0.2). In analyzing continuous variables across two groups, we determined the statistical significance of variables with normal distribution through the independent Student's t-test. Conversely, for variables not conforming to normal distribution, we employed the Mann-Whitney U test (also known as the Wilcoxon rank-sum test) to analyze differences. For categorical variables, we relied on the Chi-square test or Fisher's exact test to evaluate statistical significance between two groups. In all instances, statistical P-values were considered from a bilateral perspective, and a threshold of P < 0.05 was established to denote statistical significance. This rigorous approach ensures a comprehensive and accurate assessment of the data, adhering to the highest standards of statistical analysis." Results 2.1 Technology Roadmap 2.2 ESTIMATE score distribution between patients with sepsis and those without sepsis In our analysis, we initially employed the ESTIMATE algorithm to calculate four immune-related metrics for sepsis, including the immune score, stroma score, ESTIMATE score, and tumor purity 21 .Following this, we delved into comparing these scores between sepsis and non-sepsis patient groups. As depicted in the heatmap of Fig. 2 A, there was a notable disparity in the distribution of these four immune-related scores between sepsis patients and normal samples. Notably, sepsis patients exhibited elevated stroma scores (Anova test, p < 2.2e-16, as seen in Fig. 2 B) and tumor purity (Anova test, p = 7.9e-09, Fig. 2 C), while showing reduced ESTIMATE scores (Anova test, p = 7.9e-09, Fig. 2 D) and immune scores (Anova test, p < 2.2e-16, Fig. 2 E). This analysis provides valuable insights into the immunological landscape of sepsis, highlighting significant differences in immune and stromal components between sepsis and non-sepsis conditions. 2.3. Identification and correlation analysis of immune-related gene modules in sepsis In our study, we conducted a WGCNA on the sepsis dataset to pinpoint gene modules associated with sepsis immunity. The analysis revealed that the optimal soft threshold was 5, which achieved the lowest mean connectivity (illustrated in Fig. 3 a-A and Fig. 3 a-B, respectively). Figure 3 a-C displayed the gene clustering numbers, with various modules being differentiated by distinct colors. Subsequent steps involved identifying gene modules in relation to the four immune-related scores.The correlation heatmap between different gene modules and the four immune-related scores is shown in Fig. 3 a-D. This heatmap revealed intriguing correlations; for instance, the matrix scores exhibited the strongest association with the darkgreen module. Conversely, the immune scores were most closely linked with the grey module. We observed that both the ESTIMATE score and tumor purity showed the highest correlation with the black module. These findings offer a nuanced understanding of the relationships between specific gene modules and key immune-related scores in the context of sepsis, providing valuable insights into the genetic underpinnings of this complex condition. In our analysis of gene clustering within different color-coded modules, we observed a notable similarity in expression patterns among genes grouped in the same colored module (as depicted in Fig. 3 b-A). Further exploration into the inter-modular relationships revealed relatively low correlation levels between different modules, which is illustrated in Fig. 3 b-B. We then focused on detailing the heatmap of correlations between these diverse colored modules and sepsis. This analysis brought to light that the darkmagenta module demonstrated the most significant negative correlation with sepsis (r = -0.78), whereas the brown module exhibited the most substantial positive correlation with the disease (r = 0.7), as shown in Fig. 3 b-C.Subsequently, we presented detailed scatter plots illustrating the correlation between the brown module and its associated genes. This scatter plot analysis revealed a significant correlation (p < 0.05, shown in Fig. 3 b-D). Based on these findings, the genes within the brown module were ultimately selected as the final identified immune-related genes. This decision was grounded in the strong correlation these genes exhibited with sepsis, underscoring their potential importance in understanding the disease's immunological aspects. 2.4. Expression differences and biological pathway enrichment analysis of IRGs in sepsis To deepen our understanding of IRGs in sepsis, we first sourced a set of IRGs from the ImmPort database. These were then intersected with disease-related genes identified through WGCNA, and the intersection was visually represented in a Venn diagram (Fig. 4 A). This process led to the identification of 108 IRDEGs specifically expressed in the context of sepsis(Table 2 ). To visually compare their expression patterns, we utilized the R-package 'pheatmap' to create a heatmap. This heatmap, displayed in Fig. 4 B, clearly indicated distinct expression patterns of these 108 genes between sepsis and non-sepsis patients.Furthermore, our analysis revealed that genes significantly overexpressed in sepsis patients were predominantly enriched in biological pathways such as Osteoclast Differentiation, B Cell Receptor Signaling Pathway, Th17 Cell Differentiation, and T Cell Receptor Signaling Pathway. Conversely, genes markedly upregulated in healthy patients showed significant enrichment in the Chemokine Signaling Pathway, Th17 Cell Differentiation, JAK-STAT Signaling Pathway, PD-L1 Expression, and the PD-1 Checkpoint Pathway in Cancer, among other pathways. These findings offer crucial insights into the distinct immunological landscapes characterizing sepsis patients compared to healthy individuals." Table 2 List of sepsis-related immune genes. IL1R2 PROK2 IFNAR1 GMFG C5AR1 HSPA1A IFNAR2 TRBV7-7 NFAT5 IL18R1 TNFSF13B PLXNC1 ACVR1B IL10RB BTK MR1 TRBV6-6 NCR3 SORT1 TLR8 CYBB SLC11A1 TANK IGF1R LYN VAV1 IL21R HGF JAK2 IL1B PIK3CB IFNGR2 VIM PIK3CG TRBV5-6 TRBV5-5 MMP9 IL1RAP CHUK NFKBIA CSF2RA TRBV4-1 LTB4R IL12RB1 TRBV7-4 PLSCR1 IFNGR1 AQP9 IL10 IL18 MAP3K8 SEMA4A NFKB1 TRBV5-1 SOCS3 TLR1 NAMPT PIK3CA FPR1 SYK CSF2RB TGFB1 TRAV38-1 IL18RAP MAPK14 APOBEC3A FGR CMTM1 TRBV7-8 BCL3 TRBV9 ILK C3AR1 IL4R PDGFC HCK CKLF STAT3 PPP3CA CRLF3 B2M IL1R1 CCR1 SOS2 HSPA1B VAV3 TRAJ25 CMTM4 MAPK1 TRBV6-4 FCER1G NEDD4 TBK1 CMTM6 CXCL16 BCL10 FOS TRBV11-2 GRB2 RETN FPR2 TGFBR1 NFKBIZ CXCR1 IL32 KRAS TRAV30 NFAT5 2.5. Functional Enrichment Analysis of IRDEGs Reveals To elucidate the potential molecular mechanisms underpinning the IRDEGs, we conducted both GO and KEGG functional enrichment analyses on the 108 IRDEGs. The KEGG analysis revealed significant enrichment of these genes in pathways closely linked to cancer immunity. Notable pathways include Osteoclast Differentiation, Th17 Cell Differentiation, Cytokine-Cytokine Receptor Interaction, B Cell Receptor Signaling Pathway, T Cell Receptor Signaling Pathway, and the JAK-STAT Signaling Pathway, as illustrated in Fig. 5 A and Table 3 .Furthermore, the GO functional enrichment analysis shed light on their significant contribution to critical biological processes. These encompass a range of processes such as the Cytokine-Mediated Signaling Pathway, Positive Regulation of Cytokine Production, Leukocyte Mediated Immunity, Immune Receptor Activity, Growth Factor Receptor Binding, and the T Cell Receptor Complex, depicted in Fig. 5 B and Table 4 .These findings provide a deeper understanding of the roles played by these 108 IRDEGs, particularly in their contribution to key immune pathways and processes. 2.6 Immune-related gene features screened by machine learning algorithm and Venn diagram display To further screen for more important features in IRDEGs, we used five common Machine Learning algorithms, These include Elastic Net, LASSO regression, RF, Boruta, XGBoost decision trees. LASSO regression identified 53 important genetic features (Fig. 6 A); Elastic network identified 38 important genetic features (Fig. 6 B); RF identified 108 important genetic features (Fig. 6 C); 61 important gene features were identified by Boruta algorithm (Fig. 6 D); And XGBoost identified 20 important genetic features (Fig. 6 E). As shown in the Venn diagram (Fig. 6 F), the five machine learning algorithms collectively identified 11 IIRGs, which we finally identified as marker genes. 2.7. High-performance sepsis prediction model built with six machine learning algorithms After that, we used six different machine learning algorithms to build sepsis prediction models. The results showed that these six prediction models all had high AUC value (Fig. 7 A), and the C-index and F1-score of the models were also high (Fig. 7 B), indicating that the prediction model we built had high prediction performance. 2.8. Independent dataset validates machine learning algorithm model related to sepsis prediction Further, we validated the predictive performance of our model in independent sepsis dataset. In the GSE154918 dataset, we found that the models constructed using six different machine learning algorithms all had relatively high AUC values (Fig. 8 A,Table S2 ), all greater than 0.75, and the AUC values of the pda model were as high as 0.901. At the same time, we also found that the constructed model had a relatively low C-index and F1-score (Fig. 8 B), indicating that our model had a good and stable predictive performance for sepsis. 2.9. The importance and contribution of genetic features in different models Further, we explored the 11 IIRGs in different models. In the NB model, gene MAPK14 made the greatest contribution to sample prediction (Fig. 9 a-A). In the LogitBoost model, gene IL10 made the largest contribution to sample prediction (Fig. 9 a-B); For GBM model, gene IL21R made the largest contribution to sample prediction (Fig. 9 a-C); For cforest model, gene MAPK14 made the largest contribution to sample prediction (Fig. 9 a-D). In avNNet model, gene JAK2 made the largest contribution to sample prediction (Fig. 9 a-E). For the pda model, the gene MAPK14 contributes the most to the sample prediction (Fig. 9 a-F). In the NB model, gene JAK2 contributes the most to the model (Fig. 9 b-A). For the LogitBoost model, gene SOCS3 contributes the most to the model (Fig. 9 b-B); In the GBM model, gene NCR3 made the largest contribution to the model (Fig. 9 b-C). For cforest, gene MAPK14 made the largest contribution to the model (Fig. 9 b-D). For avNNet, gene JAK2 contributed the most in the model (Fig. 9 b-E). For pda, the gene MAPK14 contributes the most in the model (Fig. 9 b-F). 2.10. Analysis of the relationship between IIRGs and immune infiltration To further investigate the association between the final 11 IIRGs and immune infiltration in sepsis, we employed both CIBERSORT(Table S3 ) and ssGSEA ༈Table S4 ༉methodologies to assess the immune infiltration in patients from a sepsis dataset. Using CIBERSORT, we observed significant disparities in immune cell infiltration between sepsis patients and healthy samples (as depicted in Fig. 10 a-A). Specifically, immune cells like T cells CD4 memory resting, T cells CD8, NK cells resting, B cells naive, and T cells CD4 naive were predominantly found in healthy patients. Conversely, other immune cells, including Neutrophils, T cells regulatory (Tregs), Macrophages M0, and Monocytes, were notably abundant in sepsis patients. Similarly, the ssGSEA analysis revealed marked differences in immune infiltration between sepsis patients and healthy individuals (illustrated in Fig. 10 a-B). In sepsis patients, immune cells such as Neutrophils, Th2 cells, Macrophages, Mast cells, iDC, DC, Th17 cells, Treg cells, and aDC were significantly enriched. On the other hand, immune cells like NK CD56bright cells, TNK cells, Th1 cells, NK CD56dim cells, and Tfh cells were more prevalent in healthy patients. These findings offer valuable insights into the immune landscape of sepsis, highlighting the distinct immune cell profiles in sepsis patients compared to healthy individuals and underscoring the potential roles of these 11 IIRGs in modulating immune responses in sepsis. In our study, we delved into the correlation between the expression of 11 IIRGs and the immune infiltration levels in patients, subsequently visualizing these relationships through correlation circles. Within the CIBERSORT analysis, we discovered a significant correlation between the expression of these 11 IIRGs and various immune cells, notably T cell CD8, T cells CD4 memory resting, Macrophages M0, and Neutrophils (as shown in Fig. 10 b-A). Similarly, in the ssGSEA analysis, a strong correlation emerged between the expression of these genes and immune cells such as T cell CD8, B cells, Cytotoxic cells, Macrophages, NK cells, and T cells (depicted in Fig. 10 b-B). 2.11. Analysis of the relationship between IIRGs and immune checkpoints Furthermore, we explored the interplay between the expression of these 11 IIRGs and immune checkpoint genes. Initially, we compared the expression patterns of immune checkpoint genes in sepsis patients and healthy samples. This comparison revealed a notable difference in the expression of these genes between the two groups, suggesting a potential impact of immune factors on the progression of sepsis (illustrated in Fig. 11 A).Subsequent analysis focused on the correlation between the expression of these 11 IIRGs and immune checkpoint genes, again represented through correlation circles. Notably, genes such as IL21R, NCR3, and TRAV30 from our characteristic set exhibited a significant positive correlation with the expressions of most immune checkpoint genes (Pearson correlation analysis, P < 0.05), including HLA-DPB1, HLA-DPA1, HLA-DQB2, HLA-DRB1, HLA-DQA1 (shown in Fig. 11 B). Conversely, the remaining IIRGs tended to show a negative correlation with the expression of most immune checkpoint-related genes (Pearson correlation analysis, P < 0.05), as exemplified by the same set of genes (depicted in Fig. 11 B). These findings highlight complex interactions between IIRGs and immune checkpoints, offering valuable insights into their potential roles in the immune dynamics of sepsis. 2.12. Protein interaction network analysis of important immunity related genes we explored the interaction of these 11 IIRGs in the STRING database(shown in Fig. 12 )., and found that these 11 genes had strong interaction with each other, especially the genes JAK2 and IL10 had a high degree in the network, that is, they had strong interaction with other genes. 2.13. Association analysis of IIRGs and drug sensitivity In order to explore the relationship between these 11 IIRGs we finally identified and drug sensitivity, we first used the pRRophetic package to calculate the IC50 of 14 drugs corresponding to the samples in the sepsis dataset GSE134347 for CCLE database. Our subsequent results showed that among the 14 drugs, except for PD.0325901, PF2341066, PHA.665752, the IC50 of the remaining drugs was significantly different between sepsis patients and healthy patients (P < 0.05, Fig. 13 A). Indicating that there is indeed a difference in drug efficacy between sepsis patients and healthy patients. Furthermore, we also conducted correlation analysis on 11 IIRGs and drug sensitivity IC50, and the results showed that the 11 IIRGs we finally identified were indeed significantly correlated with drug sensitivity (Fig. 13 B). For example, the expression of GRB2 gene was strongly and positively correlated with the IC50 of Erlotinib (r = 0.51, P < 0.05), while it was negatively correlated with the IC50 of PD.0325901 (r = -0.61, P < 0.05). Discussion In a comprehensive study conducted across 27 academic hospitals from 2005 to 2014, there was a notable increase in septic shock incidences 22 , 23 , rising from 12.8 to 18.6 cases per 1,000 hospital admissions, even though mortality rates slightly decreased from 55–51% 24 . This uptrend is attributed to factors 25 , 26 , 27 , 28 such as aging populations, increased immunosuppression, and the prevalence of multi-drug resistant infections, emphasizing the ongoing challenge of sepsis as a critical global health concern 7 , 29 . While traditional inflammatory markers are essential in diagnosing various sepsis types, there's a notable gap in research regarding immune exhaustion in septic patients 30 , 31 , which can lead to either under-treatment or overtreatment 32 , 33 , 34 . In response to these challenges, our team has developed an innovative multi-biomarker model employing machine learning techniques. This model successfully identifies 11 key IIRGs., enhancing the ability to detect sepsis and predict drug treatment responsiveness. This advancement not only facilitates the early identification of septic patients but also aids in evaluating their immune status, thereby laying the groundwork for more tailored and precise treatment approaches. Additionally, our research highlights the potential effectiveness of immune checkpoint blockade therapy in sepsis treatment, marking a significant stride in the field. In our study, we utilized RNA-seq data from GSE154918 and identified sepsis-related disease expression genes through WGCNA. Subsequently, by intersecting these genes with immune-related genes obtained from the ImmPort database( https://www.immport.org/shared/home ), we pinpointed 108 immune-related disease expression genes linked to sepsis. Employing five different commonly used machine learning algorithms for CV, we identified 11 key IIRGs.Further, we leveraged six prevalent machine learning methods to sift through these genes, aiming to find the algorithm with the best classification performance in our validation dataset, leading to the development of a new predictive model. This model demonstrated precision in identifying septic patients, thereby validating the predictive value of these 11 IIRGs in sepsis. The genes identified are GRB2, IL21R, NCR3, TRAV30, IL10, PDGFC, APOBEC3A, MAPK14, JAK2, SOCS3 , and PLSCR1 . A comprehensive literature review revealed that six of these genes - GRB2, IL21R, NCR3, TRAV30, PDGFC , and PLSCR1 - had not been previously associated with sepsis.In-depth exploration of the roles of these genes in sepsis poses a significant challenge 35 . However, such research could potentially reveal novel therapeutic targets or disease mechanisms, offering fresh perspectives and theoretical foundations for developing new treatments for sepsis. In our constructed PPI network, IL10 , as part of a network comprising 11 key IIRGs, demonstrates a high degree of connectivity. Notably, there is a significant correlation between IL-10 and MAPK14 , along with its related signaling pathways. MAPK14 is one of the pivotal pathways regulating IL-10 36, thereby emerging as a potential target for sepsis treatment. Recent studies underscore the significance of MAPK14 in sepsis. For instance, one study revealed that inhibiting MAPK14 significantly reduces inflammatory cytokine levels, thereby improving survival rates and myocardial damage in sepsis patients, highlighting its critical role in immune responses and its potential as a therapeutic targe t​​ 37 .Moreover, another study identified MAPK14 as a key gene within a novel set of mitochondrial-related gene characteristics for diagnosing sepsis 38 . This study developed a diagnostic model utilizing six genes, including MAPK14 , emphasizing its diagnostic value in sepsis, aligning with our findings. Further research focused on key genes associated with ferroptosis in sepsis found that MAPK14 plays a crucial role in regulating ferroptosis during sepsis 39 , providing valuable insights for exploring MAPK -related ferroptosis mechanisms in sepsis.Collectively, these studies position MAPK14 as a critical focal point in understanding and managing sepsis. Our PPI network analysis also reveals that MAPK14 is closely linked with pathways involving PDGFRB , SOS2 , IL-10 , and GRB2 , further affirming the potential of MAPK14 as an attractive target for sepsis treatment. The GRB2 protein plays a pivotal role 40 in several critical biological processes, including cell growth, proliferation, metabolism, embryonic development, and the differentiation of cancer cells. It is particularly crucial in the signal transduction of immune cells 41 , with the function of Gab2 being proven to be essential in the progression of various cancers, often associated with thrombosis and inflammation. In the field of sepsis research, early studies primarily focused on comprehensive clinical and molecular analyses, aimed at identifying and prioritizing pathways that affect the survival rates of sepsis patients 42 , 43 . However, these studies did not specifically highlight the role of the GRB2 protein in these processes.Recent research has unveiled new insights. For instance, LPS or Streptococcus pneumoniae, through the activation of TAK1 by Gab2, can integrate the signaling pathways triggered by various inflammatory receptors into a common downstream pathway, leading to lung damage 44 . Additionally, studies have discovered the interaction between LRRC8A, JAK2, and GRB2 in the transformation of myofibroblasts stimulated by TGF-β1 45, 46 . Through our constructed PPI network analysis, we observed a high degree of connectivity between JAK2 and GRB2 within a network composed of 11 IIRGs.However, to date, research on the interaction between JAK2 and GRB2 has been primarily concentrated in the field of oncology, with little reporting in the context of sepsis. Therefore, delving into the interaction between GRB2 and JAK2 in sepsis could lead to groundbreaking discoveries, offering new perspectives and strategies for the treatment of sepsis. Alterations in the expression levels of key immune infiltration genes play a critical role in the onset, progression, and treatment of sepsis 47 . In light of this, we employed the CIBERSORT algorithm to analyze the relationship between the expression of these pivotal genes and immune cell infiltration in septic patients. Our results revealed a significant negative correlation between MAPK14 , JAK2 , SOCS3 , and PLSCR1 with T cell CD8 immune cells, while IL21R and NCR3 showed a significant positive correlation with T cell CD8. These findings are crucial for a deeper understanding of the immune response characteristics in septic patients.Furthermore, SOCS3, MAPK14 , and IL-10 exhibited a significant positive correlation with specific immune cells, Macrophages M0. We also analyzed these key genes and their relationship with immune infiltration in septic patients using the ssGSEA algorithm. The study found that characteristic genes like MAPK14, JAK2, SOCS3, PLSCR1, IL-10 , and PDGFC were significantly positively correlated with Macrophages, while IL21R and NCR3 showed a significant negative correlation with Macrophages, and positive correlations with various NK and T cells. This reflects their diverse roles in immune regulation.Further research into these IIRGs, especially those not previously directly linked to sepsis such as IL21R, NCR3, PLSCR1 , etc., and how they influence the progression of sepsis will be highly significant 48 . Sepsis and cancer share many common pathophysiological characteristics, with immune suppression mechanisms in both involving dysfunctions in myeloid and lymphoid cells, ultimately leading to impaired antibacterial phagocytosis and antitumor cytotoxicity 49 . Recent studies have indicated that antitumor drugs might alleviate lung damage during acute sepsis 50 , and the occurrence of sepsis might also reduce the risk of certain cancers 51 . Therefore, exploring the immune-related aspects between sepsis and cancer represents a novel area of investigation.Our study has discovered that these 11 IIGs are closely linked with immune checkpoints, which play a crucial role in regulating the immune system, especially in tumor immune evasion 52 . This suggests the existence of tumor-like immune escape mechanisms in sepsis 53 , 54 , 55 . Previous research has shown that anti- PD-L1 peptides could improve survival rates in mice infected with fungi 55 , indicating that immune checkpoint pathways might play a role in immune suppression during the later stages of sepsis.Through KEGG functional enrichment analysis, we found that pathways related to cancer immunity are significantly enriched in sepsis, suggesting a possible intersection of immune regulation between these two diseases. Surprisingly, we also discovered that the expression of these IIGs is closely associated with the sensitivity and resistance to antitumor drugs, with the GRB2 gene showing a significant positive correlation with paclitaxel 56 . While studies have found that the GRB2 gene could increase sensitivity to paclitaxel in non-small cell lung cancer, its role in enhancing sensitivity to paclitaxel through the activation of the MAPK pathway in sepsis is yet to be reported.This study not only demonstrates the efficacy of immune-related genes as biomarkers for diagnosing sepsis but also reveals their significant correlation with genes related to immune checkpoints. Modulating the expression or function of these genes could help restore normal immune responses or improve the drug responsiveness of patients with sepsis, potentially increasing survival rates. These findings might pave the way for new potential therapeutic targets in sepsis treatment, laying the groundwork for the development of more effective drug interventions for this disease. While the expression of our constructed 11 characteristic genes demonstrates good and stable predictive performance in identifying sepsis, closely correlating with patients' immune infiltration status and drug sensitivity, several limitations still need to be acknowledged and addressed in subsequent research. Firstly, our study primarily utilized the dataset from GSE154918 (n = 105) and GSE134347 (n = 298). Although the sample size of these dataset is relatively large, they might not be sufficient to represent all populations and clinical scenarios, which could limit the generalizability of our results. Secondly, we observed that models trained on one dataset (GSE154918) did not perform optimally on other independent dataset, indicating a need to collect and validate more data from diverse ethnic backgrounds.Moreover, our data analysis was based on data from public databases, which might contain recording errors or missing information. If strict quality control measures are not applied to the data, it could impact the reliability of our analysis results. Therefore, more prospective and mechanistic studies are required to further validate and refine the findings related to this research. Our study expands this field further by revealing six key molecules that have previously received limited attention: GRB2, IL21R, NCR3, TRAV30, PDGFC , and PLSCR1 . These findings offer preliminary clues for precise prediction of sepsis and improving the response of septic patients to drug treatments. Yet, the translation of these findings into clinical practice requires additional experimental and clinical research to confirm how these specific genes affect the response of septic patients to particular drugs.In summary, given the life-threatening nature of sepsis and its increasing impact, investing in biomarker research is a critical step in addressing this global healthcare challenge. Declarations Author contributions: Weichuan Xiong: analyzing data and writing the manuscript;Rui Xiao: analyzing data;Yian Zhan: designing research studies;Fangpeng Liu: acquiring data,and designing research studies. Competing interest: The authors have declared that no conflict of interest exists. Funding: None. References Rudd KE, Johnson SC, Agesa KM, Shackelford KA, Tsoi D, Kievlan DR , et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study. The Lancet 2020, 395 (10219) : 200-211. Cohen M, Banerjee D. Biomarkers in Sepsis: A Current Review of New Technologies. Journal of intensive care medicine 2023 : 08850666231194535. Z Oikonomakou M, Gkentzi D, Gogos C, Akinosoglou K. Biomarkers in pediatric sepsis: a review of recent literature. Biomarkers in Medicine 2020, 14 (10) : 895-917. Zeng Z, Peng Y, Yuan Z. Research advances of sepsis biomarkers. Zhonghua Shao Shang za zhi= Zhonghua Shaoshang Zazhi= Chinese Journal of Burns 2023, 39 (7) : 679-684. Barichello T, Generoso JS, Singer M, Dal-Pizzol F. Biomarkers for sepsis: more than just fever and leukocytosis—a narrative review. Critical care 2022, 26 (1) : 14. Yang Z, Cui X, Song Z. Predicting Sepsis Onset in ICU Using Machine Learning Models: A Systematic Review and Meta-Analysis. 2023. Goh KH, Wang L, Yeow AYK, Poh H, Li K, Yeow JJL , et al. Artificial intelligence in sepsis early prediction and diagnosis using unstructured data in healthcare. Nature communications 2021, 12 (1) : 711. Pepic I, Feldt R, Ljungström L, Torkar R, Dalevi D, Maurin Söderholm H , et al. Early detection of sepsis using artificial intelligence: a scoping review protocol. Systematic Reviews 2021, 10 (1) : 1-7. Valik JK, Ward L, Tanushi H, Johansson AF, Färnert A, Mogensen ML , et al. Predicting sepsis onset using a machine learned causal probabilistic network algorithm based on electronic health records data. Scientific reports 2023, 13 (1) : 11760. Zhang Y, Xu W, Yang P, Zhang A. Machine learning for the prediction of sepsis-related death: a systematic review and meta-analysis. BMC Medical Informatics and Decision Making 2023, 23 (1) : 283. Davis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics 2007, 23 (14) : 1846-1847. Malmstrom E, Khan HN, Veer CV, Stunnenberg M, Meijer MT, Matsumoto H , et al. The Long Non-Coding Antisense RNA JHDM1D-AS1 Regulates Inflammatory Responses in Human Monocytes. Front Cell Infect Microbiol 2022, 12: 934313. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 2008, 9: 559. Gene Ontology C. Gene Ontology Consortium: going forward. Nucleic Acids Res 2015, 43 (Database issue) : D1049-1056. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 2000, 28 (1) : 27-30. Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y , et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 2015, 12 (5) : 453-457. Barbie DA, Tamayo P, Boehm JS, Kim SY, Moody SE, Dunn IF , et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature 2009, 462 (7269) : 108-112. Szklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M , et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res 2017, 45 (D1) : D362-D368. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D , et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003, 13 (11) : 2498-2504. Chin CH, Chen SH, Wu HH, Ho CW, Ko MT, Lin CY. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014, 8 Suppl 4 (Suppl 4) : S11. Yoshihara K, Shahmoradgoli M, Martinez E, Vegesna R, Kim H, Torres-Garcia W , et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat Commun 2013, 4: 2612. Luhr R, Cao Y, Soederquist B, Cajander S. Trends in sepsis mortality over time in randomised sepsis trials: a systematic literature review and meta-analysis of mortality in the control arm, 2002–2016. Critical care 2019, 23: 1-9. Chicco D, Jurman G. Survival prediction of patients with sepsis from age, sex, and septic episode number alone. Scientific reports 2020, 10 (1) : 17156. Kadri SS, Rhee C, Strich JR, Morales MK, Hohmann S, Menchaca J , et al. Estimating ten-year trends in septic shock incidence and mortality in United States academic medical centers using clinical data. Chest 2017, 151 (2) : 278-285. Kaukonen K-M, Bailey M, Suzuki S, Pilcher D, Bellomo R. Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000-2012. Jama 2014, 311 (13) : 1308-1316. Esper AM, Martin GS. Extending international sepsis epidemiology: the impact of organ dysfunction. Critical care 2009, 13: 1-3. Blanco J, Muriel-Bombín A, Sagredo V, Taboada F, Gandía F, Tamayo L , et al. Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. Critical care 2008, 12: 1-14. Harrison DA, Welch CA, Eddleston JM. The epidemiology of severe sepsis in England, Wales and Northern Ireland, 1996 to 2004: secondary analysis of a high quality clinical database, the ICNARC Case Mix Programme Database. Critical care 2006, 10 (2) : 1-10. Wu M, Du X, Gu R, Wei J. Artificial intelligence for clinical decision support in sepsis. Frontiers in Medicine 2021, 8: 665464. Wang W, Liu C-F. Sepsis heterogeneity. World Journal of Pediatrics 2023 : 1-9. Davenport EE, Burnham KL, Radhakrishnan J, Humburg P, Hutton P, Mills TC , et al. Genomic landscape of the individual host response and outcomes in sepsis: a prospective cohort study. The Lancet Respiratory Medicine 2016, 4 (4) : 259-271. Liu D, Huang S-Y, Sun J-H, Zhang H-C, Cai Q-L, Gao C , et al. Sepsis-induced immunosuppression: mechanisms, diagnosis and current treatment options. Military Medical Research 2022, 9 (1) : 1-19. Hamers L, Kox M, Pickkers P. Sepsis-induced immunoparalysis: mechanisms, markers, and treatment options. Minerva anestesiol 2015, 81 (4) : 426-439. Hotchkiss RS, Monneret G, Payen D. Immunosuppression in sepsis: a novel understanding of the disorder and a new therapeutic approach. The Lancet infectious diseases 2013, 13 (3) : 260-268. Yu C, Huang Q. Towards more efficient and robust evaluation of sepsis treatment with deep reinforcement learning. BMC Medical Informatics and Decision Making 2023, 23 (1) : 1-10. Ge J, Yan Q, Wang Y, Cheng X, Song D, Wu C , et al. IL-10 delays the degeneration of intervertebral discs by suppressing the p38 MAPK signaling pathway. Free Radical Biology and Medicine 2020, 147: 262-270. Zhang L, Han B, Liu H, Wang J, Feng X, Sun W , et al. Circular RNA circACSL1 aggravated myocardial inflammation and myocardial injury by sponging miR-8055 and regulating MAPK14 expression. Cell Death & Disease 2021, 12 (5) : 487. Hao S, Huang M, Xu X, Wang X, Song Y, Jiang W , et al. Identification and validation of a novel mitochondrion-related gene signature for diagnosis and immune infiltration in sepsis. Frontiers in Immunology 2023, 14: 1196306. Li Z, Zhang C, Liu Y, Wang F, Zhao B, Yang J , et al. Diagnostic and Predictive Values of Ferroptosis-Related Genes in Child Sepsis. Frontiers in Immunology 2022, 13: 881914. Ijaz M, Wang F, Shahbaz M, Jiang W, Fathy AH, Nesa EU. The Role of Grb2 in Cancer and Peptides as Grb2 Antagonists. Protein and peptide letters 2018, 24 (12) : 1084-1095. Yablonski D. Bridging the Gap: Modulatory Roles of the Grb2-Family Adaptor, Gads, in Cellular and Allergic Immune Responses. Front Immunol 2019, 10: 1704. Li W, Yang J, Cao Y, Gao R, Sun H, Yang K , et al. Immunological and Prognostic Role of Hub Genes Defined Gene Signature in Septic Cardiomyopathy. 2023. Zhai J, Qi A, Zhang Y, Jiao L, Liu Y, Shou S. Bioinformatics analysis for multiple gene expression profiles in sepsis. Medical Science Monitor: International Medical Journal of Experimental and Clinical Research 2020, 26: e920818-920811. Kondreddy V, Magisetty J, Keshava S, Rao LVM, Pendurthi UR. Gab2 (Grb2-Associated Binder2) plays a crucial role in inflammatory signaling and endothelial dysfunction. Arteriosclerosis, thrombosis, and vascular biology 2021, 41 (6) : 1987-2005. Poudel S. Effects of AHCC on the Interaction between T and B Lymphocytes. Texas Southern University, 2022. Chen X, Zhang F, Hu G, Li X, Wang L, Li C , et al. LRRC8A critically regulates myofibroblast phenotypes and fibrotic remodeling following myocardial infarction. Theranostics 2022, 12 (13) : 5824. Perner A, Gordon AC, De Backer D, Dimopoulos G, Russell JA, Lipman J , et al. Sepsis: frontiers in diagnosis, resuscitation and antibiotic therapy. Intensive care medicine 2016, 42: 1958-1969. Georgescu AM, Banescu C, Azamfirei R, Hutanu A, Moldovan V, Badea I , et al. Evaluation of TNF-α genetic polymorphisms as predictors for sepsis susceptibility and progression. BMC infectious diseases 2020, 20: 1-11. Mirouse A, Vigneron C, Llitjos J-F, Chiche J-D, Mira J-P, Mokart D , et al. Sepsis and cancer: an interplay of friends and foes. American journal of respiratory and critical care medicine 2020, 202 (12) : 1625-1635. Xu L, Hu G, Xing P, Zhou M, Wang D. Paclitaxel alleviates the sepsis-induced acute kidney injury via lnc-MALAT1/miR-370-3p/HMGB1 axis. Life sciences 2020, 262: 118505. Liu Z, Mahale P, Engels EA. Sepsis and risk of cancer among elderly adults in the United States. Clinical Infectious Diseases 2019, 68 (5) : 717-724. Han Y, Liu D, Li L. PD-1/PD-L1 pathway: current researches in cancer. American journal of cancer research 2020, 10 (3) : 727. Nakamori Y, Park EJ, Shimaoka M. Immune deregulation in sepsis and septic shock: reversing immune paralysis by targeting PD-1/PD-L1 pathway. Frontiers in immunology 2021, 11: 624279. Sari MI, Ilyas S. The expression levels and concentrations of PD-1 and PD-L1 proteins in septic patients: a systematic review. Diagnostics 2022, 12 (8) : 2004. Zhang T, Yu-Jing L, Ma T. Role of regulation of PD-1 and PD-L1 expression in sepsis. Frontiers in Immunology 2023, 14: 1029438. Moghbeli M, Taghehchian N, Akhlaghipour I, Samsami Y, Maharati A. Role of forkhead box proteins in regulation of doxorubicin and paclitaxel responses in tumor cells: A comprehensive review. International Journal of Biological Macromolecules 2023 : 125995. Tables Tables 3 to 4 are available in the Supplementary Files section Additional Declarations (Not answered) Supplementary Files TableS1.Detialedinformationofimmunerelatedgenes.xls TableS2.Theperformanceofpredictivemodelinvalidationcohort.xls TableS3.TheinfiltrationofimmunecellestimatebyCIBERSORT.xls TableS4.TheinfiltrationofimmunecellestimatebyssGSEA.xls Table3.Listoftop15enrichedKEGGpathways.docx Table4.Listoftop15enrichedGOterms.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4306022","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":300256761,"identity":"25cd3f2f-b035-4e39-abe0-79c524ea3fcb","order_by":0,"name":"Fangpeng Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7ElEQVRIiWNgGAWjYFCCBIYPDAY2cmzMzAcgAgcIa2GcwVCRZszH3pZAipYzhxPn8ZwxIE6LOXuOYTNvW5oxm0TOtwc/2xjk+G4kMH4uwKPFsucNSAvQLxK52w172xiMJW8kMEvPwKPF4EaO+WOILbnbpBnbGBI33EhgY+bBrwVky+HENomcZyAt9cRp4QF6v43nDBtIS4IBQS1nnhU2zgEGMht7m5lkzzkJw5lnHjZL49VyPHljwxtgVMo3Mz+T+FFmI893PPngZ3xaGBg4DJiQFEgAMWMDXg0MDOwPGH8QUDIKRsEoGAUjHAAAph9Nx/fJwSIAAAAASUVORK5CYII=","orcid":"","institution":"The First Affiliated Hospital, Jiangxi Medical College, Nanchang University","correspondingAuthor":true,"prefix":"","firstName":"Fangpeng","middleName":"","lastName":"Liu","suffix":""},{"id":300256762,"identity":"f655c079-3407-4800-9050-a225705ad798","order_by":1,"name":"Weichuan Xiong","email":"","orcid":"https://orcid.org/0000-0003-4815-575X","institution":"The First Affiliated Hospital, Jiangxi Medical College, Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Weichuan","middleName":"","lastName":"Xiong","suffix":""},{"id":300256763,"identity":"057c87b8-b056-4f02-a52e-31b6f1e123ef","order_by":2,"name":"Rui Xiao","email":"","orcid":"","institution":"The First Affiliated Hospital, Jiangxi Medical College, Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Rui","middleName":"","lastName":"Xiao","suffix":""},{"id":300256764,"identity":"7308cec1-f424-432a-8c92-44abf8f01963","order_by":3,"name":"Yian Zhan","email":"","orcid":"","institution":"The First Affiliated Hospital, Jiangxi Medical College, Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Yian","middleName":"","lastName":"Zhan","suffix":""}],"badges":[],"createdAt":"2024-04-22 13:01:37","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4306022/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4306022/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":56112314,"identity":"017b28a6-476e-412f-832d-fdd52c54ce30","added_by":"auto","created_at":"2024-05-08 16:55:48","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1405072,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTechnology Roadmap\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWGCNA: weighted gene co-expression network analysis; cforest :Conditional Inference Random Forest;LASSO: least absolute shrinkage and selection operator; NB: naive Bayes; GBM: gradient boosting machine; avNNet: model averaged neural network; pda: penalized discriminant analysis; PPI: protein-protein interaction.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/b75735fc276c12c8f87ca8b7.png"},{"id":56113154,"identity":"742a519f-53e4-4a57-bd11-95e0328973d8","added_by":"auto","created_at":"2024-05-08 17:03:48","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":180529,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eESTIMATE analysis in the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Distribution heat map of matrix score, immune score, ESTIMATE score and tumor purity in the sepsis dataset, purple represents high score and blue represents low score;(B) Distribution of stromal scores between sepsis and non-sepsis patients; (C) distribution of tumor purity between sepsis and non-sepsis patients; (D) distribution of estimated scores between patients with sepsis and those without sepsis; (E) Distribution of immune scores between patients with sepsis and those without sepsis.\u003c/p\u003e","description":"","filename":"Figure2300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/f5df7843e539c62ca635b64b.png"},{"id":56112321,"identity":"253a49f7-593c-43a0-a402-9afe89d60ba7","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2673826,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e3a. WGCNA analysis in sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Soft threshold display in scale-free network; (B) the average connectivity of different soft thresholds in scale-free networks; (C) gene-gene clustering tree; WGCNA, weighted gene correlation network analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3b. Correlation analysis of the most relevant modules in the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Clustering network diagram between the genes of different color modules; (B) heat maps of correlations between different modules, with blue representing high correlations and purple representing low correlations; (C) heat maps of correlation between different colors and sepsis, with red representing positive correlation and blue representing negative correlation; And (D) scatter plots of correlations in the brown gene module.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/e5e35049638d9d5e5148e1f1.png"},{"id":56112318,"identity":"8ac76ded-1842-406d-afd7-e7154fbb05b5","added_by":"auto","created_at":"2024-05-08 16:55:48","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":367110,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of IRDEGs in the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Intersection diagram of disease-related genes and immune-related genes identified by WGCNA; And (B) heat maps of expression of sepsis associated immune genes between sepsis and normal patients. Blue represents high expression and purple represents low expression. IRDEGs: immune-related differentially expressed genes;WGCNA, weighted gene correlation network analysis.\u003c/p\u003e","description":"","filename":"Figure4300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/ebadc747949b5a65d7d43887.png"},{"id":56112324,"identity":"1987d7d3-b581-429e-a4a2-093da36a8a75","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":114862,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional enrichment analysis of GO and KEGG\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) KEGG functional enrichment analysis based on the sepsis dataset; And (B) GO functional enrichment analysis based on the sepsis dataset. KEGG: Kyoto Encyclopedia of Genes and Genomes; GO: Gene Ontology.\u003c/p\u003e","description":"","filename":"Figure5300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/015cb3b05851049e0948f91c.png"},{"id":56112333,"identity":"db0956dd-857c-4149-83f8-0cf6422ddae6","added_by":"auto","created_at":"2024-05-08 16:55:50","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":332771,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eScreening important immune-related features in the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) LASSO regression screening for important features; (B) elastic network screening important features; (C) RF screening important features; (D) Boruta screening for important features; (E) XGBoost algorithm filters important features. LASSO: Least absolute shrinkage and selection operator.\u003c/p\u003e","description":"","filename":"Figure6300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/b388e8f14f607139cfb7391f.png"},{"id":56113157,"identity":"3ac9b9b6-69ae-4747-8c01-85ba2a469f6d","added_by":"auto","created_at":"2024-05-08 17:03:49","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":55327,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of sepsis prediction model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) ROC curve of sepsis prediction model constructed by different machine learning algorithms; And (B) C-index and F1-score of sepsis prediction models built using different machine learning algorithms. ROC: Receiver Operating Characteristic.\u003c/p\u003e","description":"","filename":"Figure7300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/296dd9a0b24455b25ff9c6db.png"},{"id":56112326,"identity":"b2545104-fd0d-45d5-b50b-ee27e7ce8e4a","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":81139,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVerifying the sepsis prediction model in the GSE154918 dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) ROC curve of sepsis prediction model constructed by different machine learning algorithms in GSE154918 dataset; (B) C-index and F1-score of sepsis prediction models constructed using different machine learning algorithms in the GSE154918 dataset. ROC: Receiver Operating Characteristic.\u003c/p\u003e","description":"","filename":"Figure8300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/61d27b8118376c97da26d173.png"},{"id":56112329,"identity":"d4348bd3-7e55-4a5f-80a4-839836487d1e","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":548981,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e9a. Contribution of different features to predicted values in different models based on the sepsis training dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Contribution of 11 IIRGs to the predicted value in the NB model; (B) the contribution of 11 IIRGs to the predicted value in the LogiBoost model; (C) the contribution of 11 IIRGsto the predicted value in GBM model; (D) the contribution of 11 IIRGs to predicted value in cforest model; (E) the contribution of 11 IIRGs to predicted value in avNNet model; And (F) the contribution of 11 IIRGsto the predicted value in the pda model. IIRGs:important immune-related genes;NB: naive Bayes; GBM: gradient boosting machine; avNNet: model averaged neural network; pda: penalized discriminant analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;9b. Contribution of different features to the predicted value in different models based on the sepsis training dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Contribution of 11 IIRGs to the predicted value in the NB model; (B) the contribution of 11 IIRGs to the predicted value in the LogiBoost model; (C) the contribution of 11 IIRGsto the predicted value in GBM model; (D) the contribution of 11 IIRGs to predicted value in cforest model; (E) the contribution of 11 IIRGs to predicted value in avNNet model; And (F) the contribution of 11 IIRGsto the predicted value in the pda model. IIRGs:important immune-related genes;NB: naive Bayes; GBM: gradient boosting machine; avNNet: model averaged neural network; pda: penalized discriminant analysis.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/30e8fac08029336d5592f708.png"},{"id":56112330,"identity":"0ce71ed2-48a4-4e66-993d-56e2fd865e64","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":1221224,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e10a. Analysis of immune infiltration in sepsis and healthy patients in the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Distribution of immune cell infiltration between sepsis and non-sepsis patients in CIBERSORT algorithm; And (B) distribution of immune cell infiltration between sepsis and non-sepsis patients in ssGSEA algorithm, with purple representing high infiltration and blue representing low infiltration.ssGSEA, single-sample gene-set enrichment analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e10b. Correlation analysis of expression of 11 IIRGs in the sepsis dataset and immune infiltration in patients\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Correlation analysis between the expression of 11 IIRGs and immune cell infiltration in 22 of CIBERSORT algorithm; (B) Correlation analysis between the expression of 11 IIRGs and immune cell infiltration in 24 in ssGSEA algorithm. ssGSEA: single sample Gene Set Enrichment Analysis; Blue for negative correlation, purple for positive correlation. IIRGs: important immune-related genes.\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/c65a594fca1ddd33001d7e78.png"},{"id":56113155,"identity":"f736bc14-e0ca-43cd-b155-6ceaf7147ace","added_by":"auto","created_at":"2024-05-08 17:03:49","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":246826,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation analysis between IIRGs and immune checkpoint related genes in the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A)Heat maps of the expression of immune checkpoint related genes in the sepsis training dataset in patients with sepsis and patients without sepsis, with blue representing low expression and purple representing high expression. (B) The correlation between the expression of IIRGs in the sepsis training dataset and immune checkpoint-related genes, blue represents negative correlation and purple represents positive correlation. IIRGs:important immune-related genes\u003c/p\u003e","description":"","filename":"Figure11300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/63ac713a825345c7b1324da9.png"},{"id":56112323,"identity":"75f3f2b0-289b-4859-88b0-a70aa6aafb65","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":1411769,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePPI analysis of 11 IIRGs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePPI analysis of 11 IIRGs in STRING database. PPI, protein-protein interaction.\u003cstrong\u003e \u003c/strong\u003eIIRGs: important immune-related genes\u003c/p\u003e","description":"","filename":"Figure12300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/c4e2f8138b4d4be07a115c42.png"},{"id":56112332,"identity":"f125144d-b4a6-4785-8eb6-324dac018122","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":279384,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDrug sensitivity analysis of the sepsis dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Box diagram of drug IC50 distribution between sepsis patients and healthy patients, ns representing P \u0026gt;= 0.05, * representing P \u0026lt; 0.05, *** representing P \u0026lt; 0.001. (B) Correlation circle diagram of 11 IIRGs and drug IC50, purple represents positive correlation, blue represents negative correlation, * represents P \u0026lt; 0.05, the darker the color, the stronger the correlation. IIRGs:important immune-related genes.\u003c/p\u003e","description":"","filename":"Figure13300PPI.png","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/f0fa88663eff432a4080da93.png"},{"id":56115053,"identity":"e10d94e9-d59e-4ae2-82d5-37363fe3f661","added_by":"auto","created_at":"2024-05-08 17:27:56","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5206499,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/ca6a2d71-5d19-43e3-84b4-340234fa5da0.pdf"},{"id":56112315,"identity":"2d9974b3-5521-4b9a-9407-69fb85f6d8b3","added_by":"auto","created_at":"2024-05-08 16:55:48","extension":"xls","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":510464,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"TableS1.Detialedinformationofimmunerelatedgenes.xls","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/e6ed605cfcf4a00c2ecc0842.xls"},{"id":56113153,"identity":"8e71cdd8-1b30-47ad-905f-8fef8c19fd91","added_by":"auto","created_at":"2024-05-08 17:03:48","extension":"xls","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":25600,"visible":true,"origin":"","legend":"","description":"","filename":"TableS2.Theperformanceofpredictivemodelinvalidationcohort.xls","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/df4922003144e9862b59db00.xls"},{"id":56112331,"identity":"7f5c2d99-78d8-4f5a-bd2f-7d1e6045f7e7","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"xls","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":103936,"visible":true,"origin":"","legend":"","description":"","filename":"TableS3.TheinfiltrationofimmunecellestimatebyCIBERSORT.xls","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/4f7198f62f4e29a997935751.xls"},{"id":56113156,"identity":"cfbad5e4-70ab-406d-810c-d4241987db27","added_by":"auto","created_at":"2024-05-08 17:03:49","extension":"xls","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":144896,"visible":true,"origin":"","legend":"","description":"","filename":"TableS4.TheinfiltrationofimmunecellestimatebyssGSEA.xls","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/70cfe2135467c3dd0e2cf610.xls"},{"id":56112320,"identity":"e26af34c-6356-4c17-a51c-8cb613f54827","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":18690,"visible":true,"origin":"","legend":"","description":"","filename":"Table3.Listoftop15enrichedKEGGpathways.docx","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/5b0f094dfa40224cc9151d7d.docx"},{"id":56112327,"identity":"2eb39dfe-7304-4331-907e-92bb6cc6e029","added_by":"auto","created_at":"2024-05-08 16:55:49","extension":"docx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":18856,"visible":true,"origin":"","legend":"","description":"","filename":"Table4.Listoftop15enrichedGOterms.docx","url":"https://assets-eu.researchsquare.com/files/rs-4306022/v1/6caa4cf222cb8d7dcf003576.docx"}],"financialInterests":"(Not answered)","formattedTitle":"Advancing Sepsis Diagnosis and Immunotherapy Machine Learning-Driven Identification of Stable Molecular Biomarkers and Therapeutic Targets","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSepsis, a critical condition stemming from the body's maladaptive response to infection, is characterized by its high prevalence and mortality rate, exerting a profound impact on healthcare systems worldwide. A landmark study featured in The Lancet in August 2020, titled 'Global, regional, and national sepsis incidence and mortality, 1990\u0026ndash;2017: analysis for the Global Burden of Disease Study'\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, provided a comprehensive evaluation of sepsis's global impact. This investigation highlighted an alarming figure of approximately 48.9\u0026nbsp;million sepsis cases and 11\u0026nbsp;million resultant fatalities worldwide in 2017, numbers that significantly surpass previous assessments. The findings of this study brought to light the criticality of sepsis as a leading cause of death globally, presenting it as a major public health issue. It also stressed the urgency of rapid and accurate diagnosis, coupled with effective treatment strategies, as pivotal factors in improving patient survival rates.\u003c/p\u003e \u003cp\u003eSepsis manifests through a disordered immune reaction, encompassing both the innate and adaptive immune systems. This disorder often leads to an overstimulation of immune cells, causing rampant inflammation and subsequent tissue harm. Current investigations in the field of sepsis are concentrating on the identification of biomarkers, including widely recognized ones\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, like \u003cem\u003eC-reactive protein (CRP), Procalcitonin (PCT), Interleukin 6 (IL-6), pro-inflammatory cytokines\u003c/em\u003e (e.g., TNF-α, IL-1β), \u003cem\u003esoluble Triggering Receptor Expressed on Myeloid cells-1 (sTREM-1)\u003c/em\u003e, endothelial biomarkers, MicroRNA, and neutrophil \u003cem\u003eCD64\u003c/em\u003e. These biomarkers, emerging from a combination of fundamental and clinical research coupled with technological progress, have yet to provide a definitive diagnostic or prognostic capability for sepsis. Although helpful in diagnosis and prognostication, many of these markers still fall short in terms of distinctiveness or clinical applicability in the treatment of sepsis\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\"\u003c/p\u003e \u003cp\u003eAs medical science evolves swiftly, machine learning models emerge as groundbreaking approaches, potentially outperforming conventional methods in the early and ongoing detection of sepsis. This advancement could lead to timelier interventions, thereby improving patient treatment outcomes\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. These models, founded on data-driven algorithms, reduce reliance on the subjective aspects of sepsis diagnosis, which traditionally hinge on clinical discretion \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Machine learning has established itself as a crucial component in medical diagnostics. It excels in discerning intricate patterns in biological data, patterns that frequently pose challenges to human experts. This is achieved through the utilization of expansive datasets and sophisticated algorithms\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eContemporary studies in machine learning for sepsis prediction are making notable strides. One notable research \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e has put forth the SepsisFinder algorithm, showcasing its proficiency in predicting sepsis earlier compared to other models like NEWS2 and GBDT under equivalent sensitivity configurations. This finding elevates SepsisFinder as a frontrunner in early sepsis detection. Additionally, a comprehensive systematic review and meta-analysis\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e explored the capacity of machine learning to forecast sepsis-induced mortality. This analysis, incorporating diverse machine learning models, has confirmed their formidable potential in this particular domain. A focused meta-analysis\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e on the use of machine learning for sepsis onset prediction in ICU settings revealed the Random Forest (RF) model's superior accuracy and the eXtreme Gradient Boosting (XGBoost) model's exemplary predictive performance. These insights affirm the significant role of machine learning in medical practice, especially for early detection and prediction of sepsis. Our study intends to amalgamate various algorithms with extensive patient data, employing multifaceted classifier machine-learning models to refine prediction precision and reliability. We are committed to augmenting sepsis early detection and enhancing intervention efficiency at crucial immune monitoring junctures. This article delves into the utility of machine learning in deriving biomarkers from immune sources for sepsis and clarifying the potential roles of immune cells in sepsis etiology. We introduce a unique multi-classifier machine learning model for sepsis, breaking away from traditional logistic regression or LASSO regression (Least Absolute Shrinkage and Selection Operator Regression) modeling approaches. Our focus is on immune-related genes, constructing diverse predictive models to boost sepsis diagnostic accuracy. By evaluating the significance of each gene through feature importance scores or SHAP values, we analyze their mechanisms related to immune cells and their correlation with drug sensitivity. This approach is pivotal in disease diagnosis, monitoring progression, evaluating treatment efficacy, and informing clinical decisions. It holds the promise of facilitating early diagnosis, customizing treatment plans, propelling personalized medicine, and ultimately elevating patient care in various medical scenarios.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e1.1 Data Download\u003c/h2\u003e\n \u003cp\u003eUtilizing the R package GEOquery (version 2.68.0) \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e ,we accessed the GEO database to retrieve expression data from the GSE154918 (n\u0026thinsp;=\u0026thinsp;105) and GSE134347 (n\u0026thinsp;=\u0026thinsp;298) \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e datasets, both of which originated from Homo sapiens. The GSE154918 dataset comprises gene expression profiles from 56 patients clinically diagnosed with Sepsis and 49 healthy control subjects, totaling 105 samples, which were used as validation dataset in this project. This dataset provides a comprehensive comparison between the gene expression patterns in patients with Sepsis and those in healthy individuals. Similarly, the GSE134347 dataset includes gene expression data from 215 Sepsis patients and 83 healthy controls, amounting to 298 samples in total, which were used as the training dataset in this project.This dataset offers an extensive resource for analyzing the genetic underpinnings of Sepsis, contrasting the expression profiles of affected individuals with those of healthy donors.\u003c/p\u003e\n \u003cp\u003eThe data platform of dataset GSE154918 was GPL20301 Illumina HiSeq 4000 (Homo sapiens); the data platform of dataset GSE134347 is GPL17586 Affymetrix Human Transcriptome Array 2. All sample data in dataset GSE154918 and GSE134347 were removed from batch effect for follow-up analysis. See Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e for specific dataset information.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eList of datasets used in this study.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDatasets Name\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSource\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eNo. Of Sample\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGSE154918\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGEO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e298\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGSE134347\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGEO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e105\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eTo gather Immune-Related Genes (IRGs), we utilized the ImmPort database, accessible at [\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.immport.org/home\u003c/span\u003e\u003c/span\u003e]. ImmPort serves as a pivotal resource in the field of immunology, offering a platform for the aggregation, organization, and dissemination of research data. It supports the life sciences research community by enabling the archival and exchange of scientific data through advanced information technology. This database not only provides a robust repository for research data but also ensures the long-term, sustainable storage of both research and clinical data. From this database, we meticulously compiled and cross-verified a list of 1509 unique IRGs. For detailed information on these genes, refer to Supplementary Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e in our manuscript.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e1.2 Estimation\u003c/h3\u003e\n\u003cp\u003eEstimation (Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data) is an algorithm used to ESTIMATE the purity of a sample tumor. Developed and maintained by MD Anderson Cancer. A simple gene expression matrix is input to infer the content levels of immune cells, stromal cells, and tumor purity of the sample. Therefore, we input the sepsis dataset GSE134347 into the ESTIMATE algorithm to calculate the immune score, stroma score, ESTIMATE score and tumor purity of each sample, and then test the differences of these four scores in different sepsis and non-sepsis subgroups by Wilcoxon. P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1.3 WGCNA algorithm was used to identify disease-related genes in the sepsis dataset.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo isolate Sepsis-Related Genes (SRGs), our initial step involved the application of Weighted Gene Correlation Network Analysis (WGCNA)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. The purpose of WGCNA is to discern modules of co-expressed genes, elucidate the nexus between gene networks and immunity, and identify pivotal genes within these networks. We utilized the \u0026apos;pickSoftThreshold\u0026apos; function to determine the optimal soft threshold, which was found to be 5, facilitating the construction of scale-free networks based on this threshold. Subsequently, topological matrices were generated, followed by hierarchical clustering. Setting 50 as the minimum gene count for each module, we dynamically sliced and identified gene modules, and computed module Eigengenes. Module Eigengenes were then used to establish inter-module correlations and to perform further hierarchical clustering. Modules exhibiting correlations above 0.25 were amalgamated, resulting in a total of 22 distinct modules. The relationship between these modules and clinical features was analyzed using Pearson or Spearman correlation analysis.\u003c/p\u003e\n\u003cp\u003eTo acquire immune-related differentially expressed genes (IRDEGs) linked to sepsis, we intersected the SRGs identified through WGCNA with the IRGs sourced from the ImmPort database. To visualize these genes, we employed the R package \u0026apos;pheatmap\u0026apos; (version 1.0.12) to create an expression heatmap.\u003c/p\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e1.4 Screening for important immune-related genes in the sepsis dataset\u003c/h2\u003e\n \u003cp\u003eTo refine the identification of important genes within IRDEGs, we employed five prevalent Machine Learning algorithms, namely Elastic Net, LASSO regression, RF, Boruta, and XGBoost decision trees. The Elastic Net is a linear regression model that incorporates both L1 and L2 norm regularization in its training dataset. LASSO regression, based on linear regression, introduces a penalty term (lambda \u0026times; absolute value of coefficient) to reduce model overfitting and enhance generalizability. Both Elastic Net and LASSO regression were implemented using the R package \u0026apos;glmnet\u0026apos;. The outcomes of LASSO regression are visually represented through diagnostic model diagrams and variable trace plots.\u003c/p\u003e\n \u003cp\u003eRF employs ensemble learning to integrate multiple decision trees. As part of the bagging (bootstrap aggregation) algorithm within ensemble learning, it is executed via the \u0026apos;caret\u0026apos; package. RF functions by aggregating predictions from multiple trees, with the final decision derived through a majority vote. Boruta, a feature selection method gaining popularity, identifies all features correlated with the dependent variable, regardless of their impact on a specific model\u0026apos;s cost function. We applied the \u0026apos;Boruta\u0026apos; package to achieve this.\u003c/p\u003e\n \u003cp\u003eXGBoost, a gradient boosting algorithm, builds its model iteratively, each time adding a CART tree that fits the residual differences from the previous trees\u0026rsquo; predictions. This is facilitated by the \u0026apos;xgboost\u0026apos; package. Additionally, each machine learning algorithm is fine-tuned using Cross-Validation (CV) for hyperparameter optimization, ensuring model performance enhancement. To bolster robustness, we repeated the optimization ten times for each resampling, each time with a different random seed.\u003c/p\u003e\n \u003cp\u003eUltimately, to achieve stable results, genes identified by all five machine learning algorithms were consolidated as the final set of important immune-related genes(IIRGs)for our ensuing predictive model development.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e1.5 The diagnostic model of IIRGs was constructed in the sepsis dataset\u003c/h2\u003e\n \u003cp\u003eIn our pursuit to develop a sophisticated sepsis response classification model, we harnessed the potential of IIRGs by training them with an array of six prevalent machine learning algorithms. This suite included Naive Bayes (NB), Conditional Inference Random Forest(cforest), LogitBoost (an advanced form of logistic regression), Gradient Boosting Machine (GBM), Model Averaged Neural Network (avNNet), and Penalized Discriminant Analysis (pda). For all these machine learning algorithms, we meticulously employed CV as a method for hyperparameter tuning, aiming to enhance the model\u0026apos;s performance and accuracy. To assure the robustness of our models, we diligently repeated this optimization process ten times, each with a unique random seed for each iteration of resampling.\u003c/p\u003e\n \u003cp\u003eFollowing the construction of classifiers using these diverse algorithmic models, we conducted a thorough analysis through validation dataset GSE154918. This step was crucial in determining the most effective algorithm in terms of classification performance within the validation dataset. Subsequently, the algorithm demonstrating the best classification efficacy was chosen for the final assembly of our sepsis prediction model. This approach underscores our commitment to precision and reliability in developing a model adept at predicting sepsis with high accuracy.\u0026quot;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e1.6 Enrichment analysis\u003c/h2\u003e\n \u003cp\u003eThe utilization of GO analysis stands as a prevalent approach for conducting comprehensive functional enrichment studies. This analysis encompasses three key areas: Molecular Function (MF), Biological Process (BP) and Cellular Component (CC)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Additionally, the KEGG database is extensively employed for its vast repository of information on genomes, biological pathways, diseases, and pharmaceuticals \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. To delve into the potential mechanisms underlying the action of the identified crucial immune-related genes, we leveraged the \u0026apos;clusterProfiler\u0026apos; package in R (version 4.8.3). This package facilitated our detailed exploration through GO annotation analysis and KEGG pathway enrichment analysis. In our study, a False Discovery Rate (FDR) threshold of less than 0.05 was set as the benchmark for statistical significance, ensuring the reliability and relevance of our findings.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e1.7 CIBERSORT\u003c/h2\u003e\n \u003cp\u003eCIBERSORT (available at [\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cibersortx.stanford.edu/\u003c/span\u003e\u003c/span\u003e]) operates on the principle of linear support vector regression, a sophisticated statistical method used in machine learning \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. This tool, available both as an R package and a web-based application, specializes in the deconvolution of expression matrices of various human immune cell subtypes. It is particularly adept at evaluating the infiltration status of immune cells in sequenced samples, utilizing a gene expression signature set representative of 22 distinct immune cell subtypes. In our research, we employed the CIBERSORT algorithm to assess the infiltration of immune cells in a composite dataset comprising different tumor samples. We then applied the Wilcoxon test to analyze the variance in immune cell infiltration between different sepsis and non-sepsis subgroups. For our study, a P-value of less than 0.05 was set as the threshold for statistical significance, ensuring the rigor and validity of our findings.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e1.8 ssGSEA immunoinfiltration analysis\u003c/h2\u003e\n \u003cp\u003eThe single-sample gene set enrichment analysis (ssGSEA) algorithm was implemented to accurately quantify the relative abundance of each type of immune cell infiltration \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Initially, specific labels were assigned to various infiltrated immune cell types, such as Activated CD8 T cell, Gamma delta T cell, Natural killer cell, and Regulatory T cell, among other human immune cell subtypes. The ssGSEA analysis then computed enrichment scores, which served as indicators of the relative abundance of each immune cell type within individual samples. To visually represent these findings, we utilized the ggplot2 package (version 3.4.2) for graphical illustration of the distribution patterns in both sepsis and control groups. Additionally, the Wilcoxon test was employed to ascertain the differences in immune cell infiltration between sepsis and non-sepsis subgroups. In our analysis, a P-value threshold of less than 0.05 was established to denote statistical significance, thereby ensuring the robustness and validity of our results.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e1.9 Correlation analysis of IIRGs and immune infiltration in the sepsis dataset\u003c/h2\u003e\n \u003cp\u003eTo delve deeper into the underlying mechanisms of IIRGs in sepsis, our study extended its analysis to the correlation between the expression of these IIRGs and immune infiltration in sepsis patients, specifically within the sepsis dataset GSE134347. Moreover, we ventured to investigate the relationship between the expression of these IIRGs and the expression of immune checkpoints in sepsis patients. For this correlation analysis, we employed Pearson correlation analysis as our primary analytical tool. To visually represent these correlations, the \u0026apos;ggcorrplot\u0026apos; R package (version 0.1.4.1) was utilized, enabling us to create detailed and informative correlation loop diagrams. This comprehensive approach allowed us to gain a clearer understanding of the interactions and potential impact of these IIRGs in the context of sepsis and endometriosis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003e1.10 PPI Network Analysis (STRING)\u003c/h2\u003e\n \u003cp\u003eThe STRING database, renowned for its comprehensive mapping of both established and speculative Protein-Protein Interaction (PPI)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, served as a pivotal resource in our study. We utilized this database to construct a PPI network for the crucial genes we identified. The parameters for this construction were meticulously set at coefficients of 0.4, 0.7, and 0.9 to ensure optimal specificity and relevance. The data derived from the STRING database were then exported and intricately visualized using Cytoscape \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, a powerful tool for complex network analysis and visualization. Furthermore, to gain deeper insights into the central components of this network, we employed the CytoHubba plug-in \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e,for an in-depth analysis of the Hub genes within the PPI network. This comprehensive approach allowed us to unravel the intricate web of interactions among key proteins and identify pivotal genes in the context of our study.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003e1.11 Drug sensitivity analysis\u003c/h2\u003e\n \u003cp\u003eThe intricate genomic alterations in various cancers play a pivotal role in influencing clinical treatment responses, often acting as reliable biomarkers for drug efficacy. In this context, the Genomics of Drug Sensitivity in Cancer (GDSC) database (accessible at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.cancerRxgene.org\u003c/span\u003e\u003c/span\u003e) stands as the most extensive publicly available repository, offering valuable insights into drug sensitivity and molecular indicators of drug response in cancer cells. This database is instrumental in uncovering data on tumor drug responses and identifying genomic markers linked to drug sensitivity. In our research, we harnessed the capabilities of the pRRophetic algorithm. This tool enables us to predict the sensitivity of patients, categorized into different clinical variable groups, to prevalent anti-cancer drugs or small molecular compounds. This prediction is based on the analysis of expression matrices from the dataset and involves calculating IC50 values. To effectively present our findings, we utilized comparative group comparison graphs, which allow a clear and detailed visual representation of the results, thereby facilitating a better understanding of the drug response dynamics in cancer treatments.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003e1.12 Statistical analysis\u003c/h2\u003e\n \u003cp\u003eOur study\u0026apos;s data processing and statistical analysis were meticulously conducted using the R software (available at [\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.r-project.org/\u003c/span\u003e\u003c/span\u003e], version 4.0.2). In analyzing continuous variables across two groups, we determined the statistical significance of variables with normal distribution through the independent Student\u0026apos;s t-test. Conversely, for variables not conforming to normal distribution, we employed the Mann-Whitney U test (also known as the Wilcoxon rank-sum test) to analyze differences. For categorical variables, we relied on the Chi-square test or Fisher\u0026apos;s exact test to evaluate statistical significance between two groups. In all instances, statistical P-values were considered from a bilateral perspective, and a threshold of P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was established to denote statistical significance. This rigorous approach ensures a comprehensive and accurate assessment of the data, adhering to the highest standards of statistical analysis.\u0026quot;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e\n \u003cdiv id=\"Sec16\" class=\"Section4\"\u003e\u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"Results","content":"\u003ch2\u003e2.1 Technology Roadmap\u003c/h2\u003e\n\u003ch2\u003e2.2 ESTIMATE score distribution between patients with sepsis and those without sepsis\u003c/h2\u003e\n\u003cp\u003eIn our analysis, we initially employed the ESTIMATE algorithm to calculate four immune-related metrics for sepsis, including the immune score, stroma score, ESTIMATE score, and tumor purity \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.Following this, we delved into comparing these scores between sepsis and non-sepsis patient groups. As depicted in the heatmap of Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eA, there was a notable disparity in the distribution of these four immune-related scores between sepsis patients and normal samples. Notably, sepsis patients exhibited elevated stroma scores (Anova test, p\u0026thinsp;\u0026lt;\u0026thinsp;2.2e-16, as seen in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eB) and tumor purity (Anova test, p\u0026thinsp;=\u0026thinsp;7.9e-09, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eC), while showing reduced ESTIMATE scores (Anova test, p\u0026thinsp;=\u0026thinsp;7.9e-09, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eD) and immune scores (Anova test, p\u0026thinsp;\u0026lt;\u0026thinsp;2.2e-16, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eE). This analysis provides valuable insights into the immunological landscape of sepsis, highlighting significant differences in immune and stromal components between sepsis and non-sepsis conditions.\u003c/p\u003e\n\u003ch2\u003e2.3. Identification and correlation analysis of immune-related gene modules in sepsis\u003c/h2\u003e\n\u003cp\u003eIn our study, we conducted a WGCNA on the sepsis dataset to pinpoint gene modules associated with sepsis immunity. The analysis revealed that the optimal soft threshold was 5, which achieved the lowest mean connectivity (illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea-A and Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea-B, respectively). Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea-C displayed the gene clustering numbers, with various modules being differentiated by distinct colors. Subsequent steps involved identifying gene modules in relation to the four immune-related scores.The correlation heatmap between different gene modules and the four immune-related scores is shown in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea-D. This heatmap revealed intriguing correlations; for instance, the matrix scores exhibited the strongest association with the darkgreen module. Conversely, the immune scores were most closely linked with the grey module. We observed that both the ESTIMATE score and tumor purity showed the highest correlation with the black module. These findings offer a nuanced understanding of the relationships between specific gene modules and key immune-related scores in the context of sepsis, providing valuable insights into the genetic underpinnings of this complex condition.\u003c/p\u003e\n\u003cp\u003eIn our analysis of gene clustering within different color-coded modules, we observed a notable similarity in expression patterns among genes grouped in the same colored module (as depicted in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb-A). Further exploration into the inter-modular relationships revealed relatively low correlation levels between different modules, which is illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb-B. We then focused on detailing the heatmap of correlations between these diverse colored modules and sepsis. This analysis brought to light that the darkmagenta module demonstrated the most significant negative correlation with sepsis (r = -0.78), whereas the brown module exhibited the most substantial positive correlation with the disease (r\u0026thinsp;=\u0026thinsp;0.7), as shown in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb-C.Subsequently, we presented detailed scatter plots illustrating the correlation between the brown module and its associated genes. This scatter plot analysis revealed a significant correlation (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, shown in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb-D). Based on these findings, the genes within the brown module were ultimately selected as the final identified immune-related genes. This decision was grounded in the strong correlation these genes exhibited with sepsis, underscoring their potential importance in understanding the disease\u0026apos;s immunological aspects.\u003c/p\u003e\n\u003ch2\u003e2.4. Expression differences and biological pathway enrichment analysis of IRGs in sepsis\u003c/h2\u003e\n\u003cp\u003eTo deepen our understanding of IRGs in sepsis, we first sourced a set of IRGs from the ImmPort database. These were then intersected with disease-related genes identified through WGCNA, and the intersection was visually represented in a Venn diagram (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eA). This process led to the identification of 108 IRDEGs specifically expressed in the context of sepsis(Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). To visually compare their expression patterns, we utilized the R-package \u0026apos;pheatmap\u0026apos; to create a heatmap. This heatmap, displayed in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eB, clearly indicated distinct expression patterns of these 108 genes between sepsis and non-sepsis patients.Furthermore, our analysis revealed that genes significantly overexpressed in sepsis patients were predominantly enriched in biological pathways such as Osteoclast Differentiation, B Cell Receptor Signaling Pathway, Th17 Cell Differentiation, and T Cell Receptor Signaling Pathway. Conversely, genes markedly upregulated in healthy patients showed significant enrichment in the Chemokine Signaling Pathway, Th17 Cell Differentiation, JAK-STAT Signaling Pathway, PD-L1 Expression, and the PD-1 Checkpoint Pathway in Cancer, among other pathways. These findings offer crucial insights into the distinct immunological landscapes characterizing sepsis patients compared to healthy individuals.\u0026quot;\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eList of sepsis-related immune genes.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIL1R2\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePROK2\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIFNAR1\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGMFG\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eC5AR1\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eHSPA1A\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIFNAR2\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTRBV7-7\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eNFAT5\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL18R1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTNFSF13B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePLXNC1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eACVR1B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL10RB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBTK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV6-6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNCR3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSORT1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTLR8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCYBB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSLC11A1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTANK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIGF1R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLYN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVAV1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL21R\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHGF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eJAK2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL1B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePIK3CB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIFNGR2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVIM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePIK3CG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV5-6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV5-5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMMP9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL1RAP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCHUK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNFKBIA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCSF2RA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV4-1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLTB4R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL12RB1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV7-4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePLSCR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIFNGR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAQP9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMAP3K8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSEMA4A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNFKB1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV5-1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSOCS3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTLR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNAMPT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePIK3CA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFPR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSYK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCSF2RB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTGFB1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRAV38-1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL18RAP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMAPK14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAPOBEC3A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFGR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCMTM1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV7-8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBCL3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eILK\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC3AR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL4R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePDGFC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHCK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCKLF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSTAT3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePPP3CA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCRLF3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eB2M\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL1R1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCCR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSOS2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHSPA1B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVAV3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRAJ25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCMTM4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMAPK1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV6-4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFCER1G\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNEDD4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTBK1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCMTM6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCXCL16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBCL10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFOS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRBV11-2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGRB2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRETN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFPR2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTGFBR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNFKBIZ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCXCR1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIL32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKRAS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTRAV30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNFAT5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003ch2\u003e2.5. Functional Enrichment Analysis of IRDEGs Reveals\u003c/h2\u003e\n\u003cp\u003eTo elucidate the potential molecular mechanisms underpinning the IRDEGs, we conducted both GO and KEGG functional enrichment analyses on the 108 IRDEGs. The KEGG analysis revealed significant enrichment of these genes in pathways closely linked to cancer immunity. Notable pathways include Osteoclast Differentiation, Th17 Cell Differentiation, Cytokine-Cytokine Receptor Interaction, B Cell Receptor Signaling Pathway, T Cell Receptor Signaling Pathway, and the JAK-STAT Signaling Pathway, as illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eA and Table \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e.Furthermore, the GO functional enrichment analysis shed light on their significant contribution to critical biological processes. These encompass a range of processes such as the Cytokine-Mediated Signaling Pathway,\u003c/p\u003e\n\u003cp\u003ePositive Regulation of Cytokine Production, Leukocyte Mediated Immunity, Immune Receptor Activity, Growth Factor Receptor Binding, and the T Cell Receptor Complex, depicted in Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eB and Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e.These findings provide a deeper understanding of the roles played by these 108 IRDEGs, particularly in their contribution to key immune pathways and processes.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003cdiv align=\"char\" class=\"colspec\"\u003e\u003cstrong\u003e2.6 Immune-related gene features screened by machine learning algorithm and Venn diagram display\u003c/strong\u003e\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003eTo further screen for more important features in IRDEGs, we used five common Machine Learning algorithms, These include Elastic Net, LASSO regression, RF, Boruta, XGBoost decision trees. LASSO regression identified 53 important genetic features (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eA); Elastic network identified 38 important genetic features (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eB); RF identified 108 important genetic features (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eC); 61 important gene features were identified by Boruta algorithm (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eD); And XGBoost identified 20 important genetic features (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eE). As shown in the Venn diagram (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eF), the five machine learning algorithms collectively identified 11 IIRGs, which we finally identified as marker genes.\u003c/p\u003e\n\u003ch2\u003e2.7. High-performance sepsis prediction model built with six machine learning algorithms\u003c/h2\u003e\n\u003cp\u003eAfter that, we used six different machine learning algorithms to build sepsis prediction models. The results showed that these six prediction models all had high AUC value (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eA), and the C-index and F1-score of the models were also high (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eB), indicating that the prediction model we built had high prediction performance.\u003c/p\u003e\n\u003ch2\u003e2.8. Independent dataset validates machine learning algorithm model related to sepsis prediction\u003c/h2\u003e\n\u003cp\u003eFurther, we validated the predictive performance of our model in independent sepsis dataset. In the GSE154918 dataset, we found that the models constructed using six different machine learning algorithms all had relatively high AUC values (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003eA,Table \u003cspan class=\"InternalRef\"\u003eS2\u003c/span\u003e), all greater than 0.75, and the AUC values of the pda model were as high as 0.901. At the same time, we also found that the constructed model had a relatively low C-index and F1-score (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003eB), indicating that our model had a good and stable predictive performance for sepsis.\u003c/p\u003e\n\u003ch2\u003e2.9. The importance and contribution of genetic features in different models\u003c/h2\u003e\n\u003cp\u003eFurther, we explored the 11 IIRGs in different models. In the NB model, gene MAPK14 made the greatest contribution to sample prediction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003ea-A). In the LogitBoost model, gene IL10 made the largest contribution to sample prediction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003ea-B); For GBM model, gene IL21R made the largest contribution to sample prediction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003ea-C); For cforest model, gene MAPK14 made the largest contribution to sample prediction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003ea-D). In avNNet model, gene JAK2 made the largest contribution to sample prediction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003ea-E). For the pda model, the gene MAPK14 contributes the most to the sample prediction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003ea-F).\u003c/p\u003e\n\u003cp\u003eIn the NB model, gene JAK2 contributes the most to the model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003eb-A). For the LogitBoost model, gene SOCS3 contributes the most to the model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003eb-B); In the GBM model, gene NCR3 made the largest contribution to the model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003eb-C). For cforest, gene MAPK14 made the largest contribution to the model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003eb-D). For avNNet, gene JAK2 contributed the most in the model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003eb-E). For pda, the gene MAPK14 contributes the most in the model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003eb-F).\u003c/p\u003e\n\u003ch2\u003e2.10. Analysis of the relationship between IIRGs and immune infiltration\u003c/h2\u003e\n\u003cp\u003eTo further investigate the association between the final 11 IIRGs and immune infiltration in sepsis, we employed both CIBERSORT(Table \u003cspan class=\"InternalRef\"\u003eS3\u003c/span\u003e) and ssGSEA ༈Table \u003cspan class=\"InternalRef\"\u003eS4\u003c/span\u003e༉methodologies to assess the immune infiltration in patients from a sepsis dataset. Using CIBERSORT, we observed significant disparities in immune cell infiltration between sepsis patients and healthy samples (as depicted in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003ea-A). Specifically, immune cells like T cells CD4 memory resting, T cells CD8, NK cells resting, B cells naive, and T cells CD4 naive were predominantly found in healthy patients. Conversely, other immune cells, including Neutrophils, T cells regulatory (Tregs), Macrophages M0, and Monocytes, were notably abundant in sepsis patients.\u003c/p\u003e\n\u003cp\u003eSimilarly, the ssGSEA analysis revealed marked differences in immune infiltration between sepsis patients and healthy individuals (illustrated in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003ea-B). In sepsis patients, immune cells such as Neutrophils, Th2 cells, Macrophages, Mast cells, iDC, DC, Th17 cells, Treg cells, and aDC were significantly enriched. On the other hand, immune cells like NK CD56bright cells, TNK cells, Th1 cells, NK CD56dim cells, and Tfh cells were more prevalent in healthy patients. These findings offer valuable insights into the immune landscape of sepsis, highlighting the distinct immune cell profiles in sepsis patients compared to healthy individuals and underscoring the potential roles of these 11 IIRGs in modulating immune responses in sepsis.\u003c/p\u003e\n\u003cp\u003eIn our study, we delved into the correlation between the expression of 11 IIRGs and the immune infiltration levels in patients, subsequently visualizing these relationships through correlation circles. Within the CIBERSORT analysis, we discovered a significant correlation between the expression of these 11 IIRGs and various immune cells, notably T cell CD8, T cells CD4 memory resting, Macrophages M0, and Neutrophils (as shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003eb-A). Similarly, in the ssGSEA analysis, a strong correlation emerged between the expression of these genes and immune cells such as T cell CD8, B cells, Cytotoxic cells, Macrophages, NK cells, and T cells (depicted in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003eb-B).\u003c/p\u003e\n\u003ch2\u003e2.11. Analysis of the relationship between IIRGs and immune checkpoints\u003c/h2\u003e\n\u003cp\u003eFurthermore, we explored the interplay between the expression of these 11 IIRGs and immune checkpoint genes. Initially, we compared the expression patterns of immune checkpoint genes in sepsis patients and healthy samples. This comparison revealed a notable difference in the expression of these genes between the two groups, suggesting a potential impact of immune factors on the progression of sepsis (illustrated in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e11\u003c/span\u003eA).Subsequent analysis focused on the correlation between the expression of these 11 IIRGs and immune checkpoint genes, again represented through correlation circles. Notably, genes such as IL21R, NCR3, and TRAV30 from our characteristic set exhibited a significant positive correlation with the expressions of most immune checkpoint genes (Pearson correlation analysis, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), including HLA-DPB1, HLA-DPA1, HLA-DQB2, HLA-DRB1, HLA-DQA1 (shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e11\u003c/span\u003eB). Conversely, the remaining IIRGs tended to show a negative correlation with the expression of most immune checkpoint-related genes (Pearson correlation analysis, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), as exemplified by the same set of genes (depicted in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e11\u003c/span\u003eB). These findings highlight complex interactions between IIRGs and immune checkpoints, offering valuable insights into their potential roles in the immune dynamics of sepsis.\u003c/p\u003e\n\u003ch2\u003e2.12. Protein interaction network analysis of important immunity related genes\u003c/h2\u003e\n\u003cp\u003ewe explored the interaction of these 11 IIRGs in the STRING database(shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e12\u003c/span\u003e)., and found that these 11 genes had strong interaction with each other, especially the genes JAK2 and IL10 had a high degree in the network, that is, they had strong interaction with other genes.\u003c/p\u003e\n\u003ch2\u003e2.13. Association analysis of IIRGs and drug sensitivity\u003c/h2\u003e\n\u003cp\u003eIn order to explore the relationship between these 11 IIRGs we finally identified and drug sensitivity, we first used the pRRophetic package to calculate the IC50 of 14 drugs corresponding to the samples in the sepsis dataset GSE134347 for CCLE database. Our subsequent results showed that among the 14 drugs, except for PD.0325901, PF2341066, PHA.665752, the IC50 of the remaining drugs was significantly different between sepsis patients and healthy patients (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e13\u003c/span\u003eA). Indicating that there is indeed a difference in drug efficacy between sepsis patients and healthy patients. Furthermore, we also conducted correlation analysis on 11 IIRGs and drug sensitivity IC50, and the results showed that the 11 IIRGs we finally identified were indeed significantly correlated with drug sensitivity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e13\u003c/span\u003eB). For example, the expression of GRB2 gene was strongly and positively correlated with the IC50 of Erlotinib (r\u0026thinsp;=\u0026thinsp;0.51, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while it was negatively correlated with the IC50 of PD.0325901 (r = -0.61, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn a comprehensive study conducted across 27 academic hospitals from 2005 to 2014, there was a notable increase in septic shock incidences\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, rising from 12.8 to 18.6 cases per 1,000 hospital admissions, even though mortality rates slightly decreased from 55\u0026ndash;51%\u003csup\u003e24\u003c/sup\u003e. This uptrend is attributed to factors\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e such as aging populations, increased immunosuppression, and the prevalence of multi-drug resistant infections, emphasizing the ongoing challenge of sepsis as a critical global health concern\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. While traditional inflammatory markers are essential in diagnosing various sepsis types, there's a notable gap in research regarding immune exhaustion in septic patients\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, which can lead to either under-treatment or overtreatment\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn response to these challenges, our team has developed an innovative multi-biomarker model employing machine learning techniques. This model successfully identifies 11 key IIRGs., enhancing the ability to detect sepsis and predict drug treatment responsiveness. This advancement not only facilitates the early identification of septic patients but also aids in evaluating their immune status, thereby laying the groundwork for more tailored and precise treatment approaches. Additionally, our research highlights the potential effectiveness of immune checkpoint blockade therapy in sepsis treatment, marking a significant stride in the field.\u003c/p\u003e \u003cp\u003eIn our study, we utilized RNA-seq data from GSE154918 and identified sepsis-related disease expression genes through WGCNA. Subsequently, by intersecting these genes with immune-related genes obtained from the ImmPort database(\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.immport.org/shared/home\u003c/span\u003e\u003cspan address=\"https://www.immport.org/shared/home\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), we pinpointed 108 immune-related disease expression genes linked to sepsis. Employing five different commonly used machine learning algorithms for CV, we identified 11 key IIRGs.Further, we leveraged six prevalent machine learning methods to sift through these genes, aiming to find the algorithm with the best classification performance in our validation dataset, leading to the development of a new predictive model. This model demonstrated precision in identifying septic patients, thereby validating the predictive value of these 11 IIRGs in sepsis. The genes identified are \u003cem\u003eGRB2, IL21R, NCR3, TRAV30, IL10, PDGFC, APOBEC3A, MAPK14, JAK2, SOCS3\u003c/em\u003e, and \u003cem\u003ePLSCR1\u003c/em\u003e. A comprehensive literature review revealed that six of these genes - \u003cem\u003eGRB2, IL21R, NCR3, TRAV30, PDGFC\u003c/em\u003e, and \u003cem\u003ePLSCR1\u003c/em\u003e - had not been previously associated with sepsis.In-depth exploration of the roles of these genes in sepsis poses a significant challenge\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. However, such research could potentially reveal novel therapeutic targets or disease mechanisms, offering fresh perspectives and theoretical foundations for developing new treatments for sepsis.\u003c/p\u003e \u003cp\u003eIn our constructed PPI network, \u003cem\u003eIL10\u003c/em\u003e, as part of a network comprising 11 key IIRGs, demonstrates a high degree of connectivity. Notably, there is a significant correlation between \u003cem\u003eIL-10\u003c/em\u003e and \u003cem\u003eMAPK14\u003c/em\u003e, along with its related signaling pathways. \u003cem\u003eMAPK14\u003c/em\u003e is one of the pivotal pathways regulating IL-10\u003csup\u003e36,\u003c/sup\u003e thereby emerging as a potential target for sepsis treatment. Recent studies underscore the significance of \u003cem\u003eMAPK14\u003c/em\u003e in sepsis. For instance, one study revealed that inhibiting \u003cem\u003eMAPK14\u003c/em\u003e significantly reduces inflammatory cytokine levels, thereby improving survival rates and myocardial damage in sepsis patients, highlighting its critical role in immune responses and its potential as a therapeutic targe\u003csup\u003et​​\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e.Moreover, another study identified \u003cem\u003eMAPK14\u003c/em\u003e as a key gene within a novel set of mitochondrial-related gene characteristics for diagnosing sepsis\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. This study developed a diagnostic model utilizing six genes, including \u003cem\u003eMAPK14\u003c/em\u003e, emphasizing its diagnostic value in sepsis, aligning with our findings. Further research focused on key genes associated with ferroptosis in sepsis found that \u003cem\u003eMAPK14\u003c/em\u003e plays a crucial role in regulating ferroptosis during sepsis\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, providing valuable insights for exploring \u003cem\u003eMAPK\u003c/em\u003e-related ferroptosis mechanisms in sepsis.Collectively, these studies position \u003cem\u003eMAPK14\u003c/em\u003e as a critical focal point in understanding and managing sepsis. Our PPI network analysis also reveals that \u003cem\u003eMAPK14\u003c/em\u003e is closely linked with pathways involving \u003cem\u003ePDGFRB\u003c/em\u003e, \u003cem\u003eSOS2\u003c/em\u003e, \u003cem\u003eIL-10\u003c/em\u003e, and \u003cem\u003eGRB2\u003c/em\u003e, further affirming the potential of \u003cem\u003eMAPK14\u003c/em\u003e as an attractive target for sepsis treatment.\u003c/p\u003e \u003cp\u003eThe GRB2 protein plays a pivotal role\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e in several critical biological processes, including cell growth, proliferation, metabolism, embryonic development, and the differentiation of cancer cells. It is particularly crucial in the signal transduction of immune cells\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, with the function of Gab2 being proven to be essential in the progression of various cancers, often associated with thrombosis and inflammation. In the field of sepsis research, early studies primarily focused on comprehensive clinical and molecular analyses, aimed at identifying and prioritizing pathways that affect the survival rates of sepsis patients\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. However, these studies did not specifically highlight the role of the GRB2 protein in these processes.Recent research has unveiled new insights. For instance, LPS or Streptococcus pneumoniae, through the activation of TAK1 by Gab2, can integrate the signaling pathways triggered by various inflammatory receptors into a common downstream pathway, leading to lung damage \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Additionally, studies have discovered the interaction between LRRC8A, JAK2, and GRB2 in the transformation of myofibroblasts stimulated by TGF-β1\u003csup\u003e45, 46\u003c/sup\u003e. Through our constructed PPI network analysis, we observed a high degree of connectivity between JAK2 and GRB2 within a network composed of 11 IIRGs.However, to date, research on the interaction between JAK2 and GRB2 has been primarily concentrated in the field of oncology, with little reporting in the context of sepsis. Therefore, delving into the interaction between GRB2 and JAK2 in sepsis could lead to groundbreaking discoveries, offering new perspectives and strategies for the treatment of sepsis.\u003c/p\u003e \u003cp\u003eAlterations in the expression levels of key immune infiltration genes play a critical role in the onset, progression, and treatment of sepsis\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. In light of this, we employed the CIBERSORT algorithm to analyze the relationship between the expression of these pivotal genes and immune cell infiltration in septic patients. Our results revealed a significant negative correlation between \u003cem\u003eMAPK14\u003c/em\u003e, \u003cem\u003eJAK2\u003c/em\u003e, \u003cem\u003eSOCS3\u003c/em\u003e, and \u003cem\u003ePLSCR1\u003c/em\u003e with T cell CD8 immune cells, while \u003cem\u003eIL21R\u003c/em\u003e and \u003cem\u003eNCR3\u003c/em\u003e showed a significant positive correlation with T cell CD8. These findings are crucial for a deeper understanding of the immune response characteristics in septic patients.Furthermore, \u003cem\u003eSOCS3, MAPK14\u003c/em\u003e, and \u003cem\u003eIL-10\u003c/em\u003e exhibited a significant positive correlation with specific immune cells, Macrophages M0. We also analyzed these key genes and their relationship with immune infiltration in septic patients using the ssGSEA algorithm. The study found that characteristic genes like \u003cem\u003eMAPK14, JAK2, SOCS3, PLSCR1, IL-10\u003c/em\u003e, and \u003cem\u003ePDGFC\u003c/em\u003e were significantly positively correlated with Macrophages, while \u003cem\u003eIL21R\u003c/em\u003e and \u003cem\u003eNCR3\u003c/em\u003e showed a significant negative correlation with Macrophages, and positive correlations with various NK and T cells. This reflects their diverse roles in immune regulation.Further research into these IIRGs, especially those not previously directly linked to sepsis such as \u003cem\u003eIL21R, NCR3, PLSCR1\u003c/em\u003e, etc., and how they influence the progression of sepsis will be highly significant\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSepsis and cancer share many common pathophysiological characteristics, with immune suppression mechanisms in both involving dysfunctions in myeloid and lymphoid cells, ultimately leading to impaired antibacterial phagocytosis and antitumor cytotoxicity\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. Recent studies have indicated that antitumor drugs might alleviate lung damage during acute sepsis\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e, and the occurrence of sepsis might also reduce the risk of certain cancers\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. Therefore, exploring the immune-related aspects between sepsis and cancer represents a novel area of investigation.Our study has discovered that these 11 IIGs are closely linked with immune checkpoints, which play a crucial role in regulating the immune system, especially in tumor immune evasion\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. This suggests the existence of tumor-like immune escape mechanisms in sepsis\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Previous research has shown that anti-\u003cem\u003ePD-L1\u003c/em\u003e peptides could improve survival rates in mice infected with fungi\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e, indicating that immune checkpoint pathways might play a role in immune suppression during the later stages of sepsis.Through KEGG functional enrichment analysis, we found that pathways related to cancer immunity are significantly enriched in sepsis, suggesting a possible intersection of immune regulation between these two diseases. Surprisingly, we also discovered that the expression of these IIGs is closely associated with the sensitivity and resistance to antitumor drugs, with the \u003cem\u003eGRB2\u003c/em\u003e gene showing a significant positive correlation with paclitaxel\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. While studies have found that the \u003cem\u003eGRB2\u003c/em\u003e gene could increase sensitivity to paclitaxel in non-small cell lung cancer, its role in enhancing sensitivity to paclitaxel through the activation of the \u003cem\u003eMAPK\u003c/em\u003e pathway in sepsis is yet to be reported.This study not only demonstrates the efficacy of immune-related genes as biomarkers for diagnosing sepsis but also reveals their significant correlation with genes related to immune checkpoints. Modulating the expression or function of these genes could help restore normal immune responses or improve the drug responsiveness of patients with sepsis, potentially increasing survival rates. These findings might pave the way for new potential therapeutic targets in sepsis treatment, laying the groundwork for the development of more effective drug interventions for this disease.\u003c/p\u003e \u003cp\u003eWhile the expression of our constructed 11 characteristic genes demonstrates good and stable predictive performance in identifying sepsis, closely correlating with patients' immune infiltration status and drug sensitivity, several limitations still need to be acknowledged and addressed in subsequent research. Firstly, our study primarily utilized the dataset from GSE154918 (n\u0026thinsp;=\u0026thinsp;105) and GSE134347 (n\u0026thinsp;=\u0026thinsp;298). Although the sample size of these dataset is relatively large, they might not be sufficient to represent all populations and clinical scenarios, which could limit the generalizability of our results. Secondly, we observed that models trained on one dataset (GSE154918) did not perform optimally on other independent dataset, indicating a need to collect and validate more data from diverse ethnic backgrounds.Moreover, our data analysis was based on data from public databases, which might contain recording errors or missing information. If strict quality control measures are not applied to the data, it could impact the reliability of our analysis results. Therefore, more prospective and mechanistic studies are required to further validate and refine the findings related to this research.\u003c/p\u003e \u003cp\u003eOur study expands this field further by revealing six key molecules that have previously received limited attention: \u003cem\u003eGRB2, IL21R, NCR3, TRAV30, PDGFC\u003c/em\u003e, and \u003cem\u003ePLSCR1\u003c/em\u003e. These findings offer preliminary clues for precise prediction of sepsis and improving the response of septic patients to drug treatments. Yet, the translation of these findings into clinical practice requires additional experimental and clinical research to confirm how these specific genes affect the response of septic patients to particular drugs.In summary, given the life-threatening nature of sepsis and its increasing impact, investing in biomarker research is a critical step in addressing this global healthcare challenge.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor contributions:\u0026nbsp;\u003c/strong\u003eWeichuan Xiong: analyzing data and writing the manuscript;Rui Xiao: analyzing data;Yian Zhan: designing research studies;Fangpeng Liu: acquiring data,and designing research studies.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interest:\u0026nbsp;\u003c/strong\u003eThe authors have declared that no conflict of interest exists.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding: None.\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eRudd KE, Johnson SC, Agesa KM, Shackelford KA, Tsoi D, Kievlan DR\u003cem\u003e, et al.\u003c/em\u003e Global, regional, and national sepsis incidence and mortality, 1990\u0026ndash;2017: analysis for the Global Burden of Disease Study. \u003cem\u003eThe Lancet\u003c/em\u003e 2020, \u003cstrong\u003e395\u003c/strong\u003e(10219)\u003cstrong\u003e:\u003c/strong\u003e 200-211.\u003c/li\u003e\n\u003cli\u003eCohen M, Banerjee D. Biomarkers in Sepsis: A Current Review of New Technologies. \u003cem\u003eJournal of intensive care medicine\u003c/em\u003e 2023\u003cstrong\u003e:\u003c/strong\u003e 08850666231194535.\u003c/li\u003e\n\u003cli\u003eZ Oikonomakou M, Gkentzi D, Gogos C, Akinosoglou K. Biomarkers in pediatric sepsis: a review of recent literature. \u003cem\u003eBiomarkers in Medicine\u003c/em\u003e 2020, \u003cstrong\u003e14\u003c/strong\u003e(10)\u003cstrong\u003e:\u003c/strong\u003e 895-917.\u003c/li\u003e\n\u003cli\u003eZeng Z, Peng Y, Yuan Z. Research advances of sepsis biomarkers. \u003cem\u003eZhonghua Shao Shang za zhi= Zhonghua Shaoshang Zazhi= Chinese Journal of Burns\u003c/em\u003e 2023, \u003cstrong\u003e39\u003c/strong\u003e(7)\u003cstrong\u003e:\u003c/strong\u003e 679-684.\u003c/li\u003e\n\u003cli\u003eBarichello T, Generoso JS, Singer M, Dal-Pizzol F. Biomarkers for sepsis: more than just fever and leukocytosis\u0026mdash;a narrative review. \u003cem\u003eCritical care\u003c/em\u003e 2022, \u003cstrong\u003e26\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 14.\u003c/li\u003e\n\u003cli\u003eYang Z, Cui X, Song Z. Predicting Sepsis Onset in ICU Using Machine Learning Models: A Systematic Review and Meta-Analysis. 2023.\u003c/li\u003e\n\u003cli\u003eGoh KH, Wang L, Yeow AYK, Poh H, Li K, Yeow JJL\u003cem\u003e, et al.\u003c/em\u003e Artificial intelligence in sepsis early prediction and diagnosis using unstructured data in healthcare. \u003cem\u003eNature communications\u003c/em\u003e 2021, \u003cstrong\u003e12\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 711.\u003c/li\u003e\n\u003cli\u003ePepic I, Feldt R, Ljungstr\u0026ouml;m L, Torkar R, Dalevi D, Maurin S\u0026ouml;derholm H\u003cem\u003e, et al.\u003c/em\u003e Early detection of sepsis using artificial intelligence: a scoping review protocol. \u003cem\u003eSystematic Reviews\u003c/em\u003e 2021, \u003cstrong\u003e10\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 1-7.\u003c/li\u003e\n\u003cli\u003eValik JK, Ward L, Tanushi H, Johansson AF, F\u0026auml;rnert A, Mogensen ML\u003cem\u003e, et al.\u003c/em\u003e Predicting sepsis onset using a machine learned causal probabilistic network algorithm based on electronic health records data. \u003cem\u003eScientific reports\u003c/em\u003e 2023, \u003cstrong\u003e13\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 11760.\u003c/li\u003e\n\u003cli\u003eZhang Y, Xu W, Yang P, Zhang A. Machine learning for the prediction of sepsis-related death: a systematic review and meta-analysis. \u003cem\u003eBMC Medical Informatics and Decision Making\u003c/em\u003e 2023, \u003cstrong\u003e23\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 283.\u003c/li\u003e\n\u003cli\u003eDavis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. \u003cem\u003eBioinformatics\u003c/em\u003e 2007, \u003cstrong\u003e23\u003c/strong\u003e(14)\u003cstrong\u003e:\u003c/strong\u003e 1846-1847.\u003c/li\u003e\n\u003cli\u003eMalmstrom E, Khan HN, Veer CV, Stunnenberg M, Meijer MT, Matsumoto H\u003cem\u003e, et al.\u003c/em\u003e The Long Non-Coding Antisense RNA JHDM1D-AS1 Regulates Inflammatory Responses in Human Monocytes. \u003cem\u003eFront Cell Infect Microbiol\u003c/em\u003e 2022, \u003cstrong\u003e12:\u003c/strong\u003e 934313.\u003c/li\u003e\n\u003cli\u003eLangfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. \u003cem\u003eBMC Bioinformatics\u003c/em\u003e 2008, \u003cstrong\u003e9:\u003c/strong\u003e 559.\u003c/li\u003e\n\u003cli\u003eGene Ontology C. Gene Ontology Consortium: going forward. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 2015, \u003cstrong\u003e43\u003c/strong\u003e(Database issue)\u003cstrong\u003e:\u003c/strong\u003e D1049-1056.\u003c/li\u003e\n\u003cli\u003eKanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 2000, \u003cstrong\u003e28\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 27-30.\u003c/li\u003e\n\u003cli\u003eNewman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y\u003cem\u003e, et al.\u003c/em\u003e Robust enumeration of cell subsets from tissue expression profiles. \u003cem\u003eNat Methods\u003c/em\u003e 2015, \u003cstrong\u003e12\u003c/strong\u003e(5)\u003cstrong\u003e:\u003c/strong\u003e 453-457.\u003c/li\u003e\n\u003cli\u003eBarbie DA, Tamayo P, Boehm JS, Kim SY, Moody SE, Dunn IF\u003cem\u003e, et al.\u003c/em\u003e Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. \u003cem\u003eNature\u003c/em\u003e 2009, \u003cstrong\u003e462\u003c/strong\u003e(7269)\u003cstrong\u003e:\u003c/strong\u003e 108-112.\u003c/li\u003e\n\u003cli\u003eSzklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M\u003cem\u003e, et al.\u003c/em\u003e The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 2017, \u003cstrong\u003e45\u003c/strong\u003e(D1)\u003cstrong\u003e:\u003c/strong\u003e D362-D368.\u003c/li\u003e\n\u003cli\u003eShannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D\u003cem\u003e, et al.\u003c/em\u003e Cytoscape: a software environment for integrated models of biomolecular interaction networks. \u003cem\u003eGenome Res\u003c/em\u003e 2003, \u003cstrong\u003e13\u003c/strong\u003e(11)\u003cstrong\u003e:\u003c/strong\u003e 2498-2504.\u003c/li\u003e\n\u003cli\u003eChin CH, Chen SH, Wu HH, Ho CW, Ko MT, Lin CY. cytoHubba: identifying hub objects and sub-networks from complex interactome. \u003cem\u003eBMC Syst Biol\u003c/em\u003e 2014, \u003cstrong\u003e8 Suppl 4\u003c/strong\u003e(Suppl 4)\u003cstrong\u003e:\u003c/strong\u003e S11.\u003c/li\u003e\n\u003cli\u003eYoshihara K, Shahmoradgoli M, Martinez E, Vegesna R, Kim H, Torres-Garcia W\u003cem\u003e, et al.\u003c/em\u003e Inferring tumour purity and stromal and immune cell admixture from expression data. \u003cem\u003eNat Commun\u003c/em\u003e 2013, \u003cstrong\u003e4:\u003c/strong\u003e 2612.\u003c/li\u003e\n\u003cli\u003eLuhr R, Cao Y, Soederquist B, Cajander S. Trends in sepsis mortality over time in randomised sepsis trials: a systematic literature review and meta-analysis of mortality in the control arm, 2002\u0026ndash;2016. \u003cem\u003eCritical care\u003c/em\u003e 2019, \u003cstrong\u003e23:\u003c/strong\u003e 1-9.\u003c/li\u003e\n\u003cli\u003eChicco D, Jurman G. Survival prediction of patients with sepsis from age, sex, and septic episode number alone. \u003cem\u003eScientific reports\u003c/em\u003e 2020, \u003cstrong\u003e10\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 17156.\u003c/li\u003e\n\u003cli\u003eKadri SS, Rhee C, Strich JR, Morales MK, Hohmann S, Menchaca J\u003cem\u003e, et al.\u003c/em\u003e Estimating ten-year trends in septic shock incidence and mortality in United States academic medical centers using clinical data. \u003cem\u003eChest\u003c/em\u003e 2017, \u003cstrong\u003e151\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e 278-285.\u003c/li\u003e\n\u003cli\u003eKaukonen K-M, Bailey M, Suzuki S, Pilcher D, Bellomo R. Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000-2012. \u003cem\u003eJama\u003c/em\u003e 2014, \u003cstrong\u003e311\u003c/strong\u003e(13)\u003cstrong\u003e:\u003c/strong\u003e 1308-1316.\u003c/li\u003e\n\u003cli\u003eEsper AM, Martin GS. Extending international sepsis epidemiology: the impact of organ dysfunction. \u003cem\u003eCritical care\u003c/em\u003e 2009, \u003cstrong\u003e13:\u003c/strong\u003e 1-3.\u003c/li\u003e\n\u003cli\u003eBlanco J, Muriel-Bomb\u0026iacute;n A, Sagredo V, Taboada F, Gand\u0026iacute;a F, Tamayo L\u003cem\u003e, et al.\u003c/em\u003e Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. \u003cem\u003eCritical care\u003c/em\u003e 2008, \u003cstrong\u003e12:\u003c/strong\u003e 1-14.\u003c/li\u003e\n\u003cli\u003eHarrison DA, Welch CA, Eddleston JM. The epidemiology of severe sepsis in England, Wales and Northern Ireland, 1996 to 2004: secondary analysis of a high quality clinical database, the ICNARC Case Mix Programme Database. \u003cem\u003eCritical care\u003c/em\u003e 2006, \u003cstrong\u003e10\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e 1-10.\u003c/li\u003e\n\u003cli\u003eWu M, Du X, Gu R, Wei J. Artificial intelligence for clinical decision support in sepsis. \u003cem\u003eFrontiers in Medicine\u003c/em\u003e 2021, \u003cstrong\u003e8:\u003c/strong\u003e 665464.\u003c/li\u003e\n\u003cli\u003eWang W, Liu C-F. Sepsis heterogeneity. \u003cem\u003eWorld Journal of Pediatrics\u003c/em\u003e 2023\u003cstrong\u003e:\u003c/strong\u003e 1-9.\u003c/li\u003e\n\u003cli\u003eDavenport EE, Burnham KL, Radhakrishnan J, Humburg P, Hutton P, Mills TC\u003cem\u003e, et al.\u003c/em\u003e Genomic landscape of the individual host response and outcomes in sepsis: a prospective cohort study. \u003cem\u003eThe Lancet Respiratory Medicine\u003c/em\u003e 2016, \u003cstrong\u003e4\u003c/strong\u003e(4)\u003cstrong\u003e:\u003c/strong\u003e 259-271.\u003c/li\u003e\n\u003cli\u003eLiu D, Huang S-Y, Sun J-H, Zhang H-C, Cai Q-L, Gao C\u003cem\u003e, et al.\u003c/em\u003e Sepsis-induced immunosuppression: mechanisms, diagnosis and current treatment options. \u003cem\u003eMilitary Medical Research\u003c/em\u003e 2022, \u003cstrong\u003e9\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 1-19.\u003c/li\u003e\n\u003cli\u003eHamers L, Kox M, Pickkers P. Sepsis-induced immunoparalysis: mechanisms, markers, and treatment options. \u003cem\u003eMinerva anestesiol\u003c/em\u003e 2015, \u003cstrong\u003e81\u003c/strong\u003e(4)\u003cstrong\u003e:\u003c/strong\u003e 426-439.\u003c/li\u003e\n\u003cli\u003eHotchkiss RS, Monneret G, Payen D. Immunosuppression in sepsis: a novel understanding of the disorder and a new therapeutic approach. \u003cem\u003eThe Lancet infectious diseases\u003c/em\u003e 2013, \u003cstrong\u003e13\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 260-268.\u003c/li\u003e\n\u003cli\u003eYu C, Huang Q. Towards more efficient and robust evaluation of sepsis treatment with deep reinforcement learning. \u003cem\u003eBMC Medical Informatics and Decision Making\u003c/em\u003e 2023, \u003cstrong\u003e23\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 1-10.\u003c/li\u003e\n\u003cli\u003eGe J, Yan Q, Wang Y, Cheng X, Song D, Wu C\u003cem\u003e, et al.\u003c/em\u003e IL-10 delays the degeneration of intervertebral discs by suppressing the p38 MAPK signaling pathway. \u003cem\u003eFree Radical Biology and Medicine\u003c/em\u003e 2020, \u003cstrong\u003e147:\u003c/strong\u003e 262-270.\u003c/li\u003e\n\u003cli\u003eZhang L, Han B, Liu H, Wang J, Feng X, Sun W\u003cem\u003e, et al.\u003c/em\u003e Circular RNA circACSL1 aggravated myocardial inflammation and myocardial injury by sponging miR-8055 and regulating MAPK14 expression. \u003cem\u003eCell Death \u0026amp; Disease\u003c/em\u003e 2021, \u003cstrong\u003e12\u003c/strong\u003e(5)\u003cstrong\u003e:\u003c/strong\u003e 487.\u003c/li\u003e\n\u003cli\u003eHao S, Huang M, Xu X, Wang X, Song Y, Jiang W\u003cem\u003e, et al.\u003c/em\u003e Identification and validation of a novel mitochondrion-related gene signature for diagnosis and immune infiltration in sepsis. \u003cem\u003eFrontiers in Immunology\u003c/em\u003e 2023, \u003cstrong\u003e14:\u003c/strong\u003e 1196306.\u003c/li\u003e\n\u003cli\u003eLi Z, Zhang C, Liu Y, Wang F, Zhao B, Yang J\u003cem\u003e, et al.\u003c/em\u003e Diagnostic and Predictive Values of Ferroptosis-Related Genes in Child Sepsis. \u003cem\u003eFrontiers in Immunology\u003c/em\u003e 2022, \u003cstrong\u003e13:\u003c/strong\u003e 881914.\u003c/li\u003e\n\u003cli\u003eIjaz M, Wang F, Shahbaz M, Jiang W, Fathy AH, Nesa EU. The Role of Grb2 in Cancer and Peptides as Grb2 Antagonists. \u003cem\u003eProtein and peptide letters\u003c/em\u003e 2018, \u003cstrong\u003e24\u003c/strong\u003e(12)\u003cstrong\u003e:\u003c/strong\u003e 1084-1095.\u003c/li\u003e\n\u003cli\u003eYablonski D. Bridging the Gap: Modulatory Roles of the Grb2-Family Adaptor, Gads, in Cellular and Allergic Immune Responses. \u003cem\u003eFront Immunol\u003c/em\u003e 2019, \u003cstrong\u003e10:\u003c/strong\u003e 1704.\u003c/li\u003e\n\u003cli\u003eLi W, Yang J, Cao Y, Gao R, Sun H, Yang K\u003cem\u003e, et al.\u003c/em\u003e Immunological and Prognostic Role of Hub Genes Defined Gene Signature in Septic Cardiomyopathy. 2023.\u003c/li\u003e\n\u003cli\u003eZhai J, Qi A, Zhang Y, Jiao L, Liu Y, Shou S. Bioinformatics analysis for multiple gene expression profiles in sepsis. \u003cem\u003eMedical Science Monitor: International Medical Journal of Experimental and Clinical Research\u003c/em\u003e 2020, \u003cstrong\u003e26:\u003c/strong\u003e e920818-920811.\u003c/li\u003e\n\u003cli\u003eKondreddy V, Magisetty J, Keshava S, Rao LVM, Pendurthi UR. Gab2 (Grb2-Associated Binder2) plays a crucial role in inflammatory signaling and endothelial dysfunction. \u003cem\u003eArteriosclerosis, thrombosis, and vascular biology\u003c/em\u003e 2021, \u003cstrong\u003e41\u003c/strong\u003e(6)\u003cstrong\u003e:\u003c/strong\u003e 1987-2005.\u003c/li\u003e\n\u003cli\u003ePoudel S. Effects of AHCC on the Interaction between T and B Lymphocytes. Texas Southern University, 2022.\u003c/li\u003e\n\u003cli\u003eChen X, Zhang F, Hu G, Li X, Wang L, Li C\u003cem\u003e, et al.\u003c/em\u003e LRRC8A critically regulates myofibroblast phenotypes and fibrotic remodeling following myocardial infarction. \u003cem\u003eTheranostics\u003c/em\u003e 2022, \u003cstrong\u003e12\u003c/strong\u003e(13)\u003cstrong\u003e:\u003c/strong\u003e 5824.\u003c/li\u003e\n\u003cli\u003ePerner A, Gordon AC, De Backer D, Dimopoulos G, Russell JA, Lipman J\u003cem\u003e, et al.\u003c/em\u003e Sepsis: frontiers in diagnosis, resuscitation and antibiotic therapy. \u003cem\u003eIntensive care medicine\u003c/em\u003e 2016, \u003cstrong\u003e42:\u003c/strong\u003e 1958-1969.\u003c/li\u003e\n\u003cli\u003eGeorgescu AM, Banescu C, Azamfirei R, Hutanu A, Moldovan V, Badea I\u003cem\u003e, et al.\u003c/em\u003e Evaluation of TNF-\u0026alpha; genetic polymorphisms as predictors for sepsis susceptibility and progression. \u003cem\u003eBMC infectious diseases\u003c/em\u003e 2020, \u003cstrong\u003e20:\u003c/strong\u003e 1-11.\u003c/li\u003e\n\u003cli\u003eMirouse A, Vigneron C, Llitjos J-F, Chiche J-D, Mira J-P, Mokart D\u003cem\u003e, et al.\u003c/em\u003e Sepsis and cancer: an interplay of friends and foes. \u003cem\u003eAmerican journal of respiratory and critical care medicine\u003c/em\u003e 2020, \u003cstrong\u003e202\u003c/strong\u003e(12)\u003cstrong\u003e:\u003c/strong\u003e 1625-1635.\u003c/li\u003e\n\u003cli\u003eXu L, Hu G, Xing P, Zhou M, Wang D. Paclitaxel alleviates the sepsis-induced acute kidney injury via lnc-MALAT1/miR-370-3p/HMGB1 axis. \u003cem\u003eLife sciences\u003c/em\u003e 2020, \u003cstrong\u003e262:\u003c/strong\u003e 118505.\u003c/li\u003e\n\u003cli\u003eLiu Z, Mahale P, Engels EA. Sepsis and risk of cancer among elderly adults in the United States. \u003cem\u003eClinical Infectious Diseases\u003c/em\u003e 2019, \u003cstrong\u003e68\u003c/strong\u003e(5)\u003cstrong\u003e:\u003c/strong\u003e 717-724.\u003c/li\u003e\n\u003cli\u003eHan Y, Liu D, Li L. PD-1/PD-L1 pathway: current researches in cancer. \u003cem\u003eAmerican journal of cancer research\u003c/em\u003e 2020, \u003cstrong\u003e10\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 727.\u003c/li\u003e\n\u003cli\u003eNakamori Y, Park EJ, Shimaoka M. Immune deregulation in sepsis and septic shock: reversing immune paralysis by targeting PD-1/PD-L1 pathway. \u003cem\u003eFrontiers in immunology\u003c/em\u003e 2021, \u003cstrong\u003e11:\u003c/strong\u003e 624279.\u003c/li\u003e\n\u003cli\u003eSari MI, Ilyas S. The expression levels and concentrations of PD-1 and PD-L1 proteins in septic patients: a systematic review. \u003cem\u003eDiagnostics\u003c/em\u003e 2022, \u003cstrong\u003e12\u003c/strong\u003e(8)\u003cstrong\u003e:\u003c/strong\u003e 2004.\u003c/li\u003e\n\u003cli\u003eZhang T, Yu-Jing L, Ma T. Role of regulation of PD-1 and PD-L1 expression in sepsis. \u003cem\u003eFrontiers in Immunology\u003c/em\u003e 2023, \u003cstrong\u003e14:\u003c/strong\u003e 1029438.\u003c/li\u003e\n\u003cli\u003eMoghbeli M, Taghehchian N, Akhlaghipour I, Samsami Y, Maharati A. Role of forkhead box proteins in regulation of doxorubicin and paclitaxel responses in tumor cells: A comprehensive review. \u003cem\u003eInternational Journal of Biological Macromolecules\u003c/em\u003e 2023\u003cstrong\u003e:\u003c/strong\u003e 125995.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 3 to 4 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4306022/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4306022/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"This research presents a novel integrated approach combining genomic analysis and machine learning to identify biomarkers and drug sensitivities specific to sepsis, aiming to facilitate personalized treatment strategies. We comprehensively examined gene expression profiles from sepsis patients and healthy controls by utilizing the Gene Expression Omnibus (GEO) database, specifically datasets GSE154918 and GSE134347. Through the application of the ESTIMATE algorithm, weighted gene co-expression network analysis (WGCNA), and a range of machine learning techniques, we identified crucial Sepsis-Related Genes (SRGs), Immune-Related Differentially Expressed Genes (IRDEGs), and Important Immune-related genes (IIRGs). Our analysis revealed significant differences in immune and stromal scores between sepsis patients and controls, highlighting the altered immune landscape in sepsis. The study also uncovered specific genes associated with drug sensitivity, providing insights into potential therapeutic targets. The predictive model developed demonstrated high accuracy in sepsis diagnosis and prognosis, validated by independent datasets. These findings contribute to understanding sepsis at a molecular level and open new avenues for developing personalized therapeutic interventions, underscoring the potential of integrating genomic analysis and machine learning in sepsis research.","manuscriptTitle":"Advancing Sepsis Diagnosis and Immunotherapy Machine Learning-Driven Identification of Stable Molecular Biomarkers and Therapeutic Targets","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-08 16:55:43","doi":"10.21203/rs.3.rs-4306022/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a6c0c608-7430-4de9-98f7-ed2c82f36613","owner":[],"postedDate":"May 8th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":31681409,"name":"Biological sciences/Biotechnology/Genomics/Phylogenomics"},{"id":31681410,"name":"Biological sciences/Drug discovery/Biomarkers/Diagnostic markers"},{"id":31681411,"name":"Biological sciences/Molecular biology/Proteomics/Protein\u0026#x2013;protein interaction networks"}],"tags":[],"updatedAt":"2024-05-08T17:03:46+00:00","versionOfRecord":[],"versionCreatedAt":"2024-05-08 16:55:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4306022","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4306022","identity":"rs-4306022","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}