Gene coexpression network approach to develop an immune prognostic model for pancreatic adenocarcinoma.
Ontology highlight
ABSTRACT: Background
Pancreatic adenocarcinoma (PAAD) is a nonimmunogenic tumor, and very little is known about the relationship between the host immune response and patient survival. We aimed to develop an immune prognostic model (IPM) and analyze its relevance to the tumor immune profiles of patients with PAAD.Methods
We investigated differentially expressed genes between tumor and normal tissues in the TCGA PAAD cohort. Immune-related genes were screened from highly variably expressed genes with weighted gene correlation network analysis (WGCNA) to construct an IPM. Then, the influence of IPM on the PAAD immune profile was comprehensively analyzed.Results
A total of 4902 genes highly variably expressed among primary tumors were used to construct a weighted gene coexpression network. One hundred seventy-five hub genes in the immune-related module were used for machine learning. Then, we established an IPM with four core genes (FCGR2B, IL10RA, and HLA-DRA) to evaluate the prognosis. The risk score predicted by IPM was an independent prognostic factor and had a high predictive value for the prognosis of patients with PAAD. Moreover, we found that the patients in the low-risk group had higher cytolytic activity and lower innate anti-PD-1 resistance (IPRES) signatures than patients in the high-risk group.Conclusions
Unlike the traditional methods that use immune-related genes listed in public databases to screen prognostic genes, we constructed an IPM through WGCNA to predict the prognosis of PAAD patients. In addition, an IPM prediction of low risk indicated enhanced immune activity and a decreased anti-PD-1 therapeutic response.
SUBMITTER: Gu X
PROVIDER: S-EPMC8042890 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Purpose We aimed to identify new prognostic biomarkers of lung adenocarcinoma (LUAD) based on cancer stem cell theory. Materials and Methods: RNA-seq and microarray data were obtained with clinical information downloaded from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) databases. Weighted gene coexpression network analysis (WGCNA) was applied to identify significant module and hub genes. The hub genes were validated via microarray data from GEO, and a prognostic signature with prognostic hub genes was constructed. Results LUAD patients enrolled from TCGA had a higher mRNA expression-based stemness index (mRNAsi) in tumor tissue than in adjacent normal tissue. Some clinical features and prognoses were found to be highly correlated with mRNAsi. WGCNA found that the green module and blue module were the most significant modules related to mRNAsi; 50 key genes were identified in the green module and were enriched mostly in the cell cycle, chromosome segregation, chromosomal region and microtubule binding. Six hub genes were revealed through the protein-protein interaction (PPI) network and Molecular Complex Detection (MCODE) plugin of Cytoscape software. Based on external verification with the GEO database, these six genes are not only expressed at different levels in LUAD and normal tissues but also associated with different clinical features. In addition, the construction of a prognostic signature with three hub genes showed high predictive value. Conclusion mRNAsi-related biomarkers may suggest a new potential treatment strategy for LUAD.
| S-EPMC7192063 | biostudies-literature
Project description:Background and aimsPancreatic adenocarcinoma (PAAD) is highly aggressive and characterized by a poor prognosis. Oxidative stress has great impacts on the occurrence and development of tumors. However, the predictive role of oxidative stress related genes on PAAD patients' prognosis remains unclear. In this study, we aimed to construct a prognostic model for PAAD based on oxidative stress genes and to evaluate its predictive value.MethodsThe Cancer Genome Atlas (TCGA) and three Gene Expression Omnibus (GEO) datasets were used to identify differentially expressed oxidative stress genes. Univariate Cox regression, Kaplan-Meier and multivariate Cox regression analysis were used to select genes and to construct a prognosis model. According to the median value of the model's risk score, patients were divided into high and low risk groups, and gene set enrichment analysis (GSEA), immune infiltration and immunotherapy effect, drug resistance and the expression of immune checkpoint related genes and synthetic driver genes of T cell proliferation were analyzed. Finally, the mRNA and protein levels of four genes in PAAD were verified by the clinical proteomic tumor analysis consortium (CPTAC) database and the immunostaining of patients' tissue.Results55 differentially expressed oxidative stress genes were identified, and four genes including MET, FYN, CTTN and CDK1 were selected to construct a prognosis model. GESA indicated that immune related pathways, metabolic pathways and DNA repair pathways were significantly enriched in the high risk group as compared to the low risk group. The frequency of genetic mutations was also significantly higher in high risk groups than that in low risk groups. Moreover, the infiltration level of 23 immune cells as well as the expression of immune checkpoint related and synthetic driver genes of T cell proliferation were significantly altered, with the better immunotherapy effect occurring in low risk group. In patient PAAD tissues, the mRNA and protein levels of these four genes were up-regulated.ConclusionWe have successfully constructed a four oxidative stress gene prognostic model that has important predictive value for PAAD patients, and this model might be a promising guidance for prognostic prediction and efficacy monitoring in clinical individualized therapy.
| S-EPMC9880292 | biostudies-literature
Project description:RNA methylation modification is a key process in epigenetics that regulates posttranscriptional gene expression. With advances in next-generation sequencing technology, 5-methylcytosine (m5C) modification has also been found in multiple RNAs. Long non-coding RNAs (lncRNAs) were proved to have a key role in cancer progression and closely related to the tumor immune microenvironment. Thus, based on the PDAC patients' clinical information and genetic transcriptome data from the TCGA database, we performed a detailed bioinformatic analysis to establish a m5C-related lncRNA prognostic risk model for PDAC patients and discovered the relationship between the risk model and PDAC immune microenvironment. Pearson correlation coefficient analysis was applied to conduct a m5C regulatory gene and m5C-related lncRNA co-expression network. Expression of m5C-related lncRNAs screened by univariate regression analysis with prognostic value showed a significant difference between pancreatic cancer and normal tissues. The least absolute shrinkage and selection operator (LASSO) Cox regression method was applied to determine an 8-m5C-related lncRNA prognostic risk model. We used principal component analysis to indicate that the risk model could distinguish all the samples clearly. The clinical nomogram also accurately predicted 1-, 1.5-, 2-, and 3-year survival time among PDAC patients. Additionally, this risk model was validated in the entire group and sub-test groups using KM analysis and ROC analysis. Combined with the clinical characteristics, the risk score was found to be an independent factor for predicting the survival of PDAC patients. Furthermore, the association between the risk model and tumor immune microenvironment was evaluated via the ESTIMATE R package and CIBERSORT method. Consequently, the results indicated that immune cells were associated with m5C-related lncRNA risk model scores and had different distribution in the high- and low-risk groups. Based on all these analyses, the m5C-related lncRNA risk model could be a reliable prognostic tool and therapeutic target for PDAC patients.
| S-EPMC8692582 | biostudies-literature
Project description:BackgroundPancreatic adenocarcinoma (PAAD) is a highly lethal and aggressive tumor with poor prognoses. The predictive capability of immune-related genes (IRGs) in PAAD has yet to be explored. We aimed to explore prognostic-related immune genes and develop a prediction model for indicating prognosis in PAAD.MethodsThe messenger (m)RNA expression profiles acquired from public databases were comprehensively integrated and differentially expressed genes were identified. Univariate analysis was utilized to identify IRGs that related to overall survival. Whereafter, a multigene signature in the Cancer Genome Atlas cohort was established based on the least absolute shrinkage and selection operator (LASSO) Cox regression analysis. Moreover, a transcription factors regulatory network was constructed to reveal potential molecular processes in PAAD. PAAD datasets downloaded from the Gene Expression Omnibus database were applied for the validations. Finally, correlation analysis between the prognostic model and immunocyte infiltration was investigated.ResultsTotally, 446 differentially expressed immune-related genes were screened in PAAD tissues and normal tissues, of which 43 IRGs were significantly related to the overall survival of PAAD patients. An immune-based prognostic model was developed, which contained eight IRGs. Univariate and multivariate Cox regression revealed that the risk score model was an independent prognostic indicator in PAAD (HR > 1, P < 0.001). Besides, the sensitivity of the model was evaluated by the receiver operating characteristic curve analysis. Finally, immunocyte infiltration analysis revealed that the eight-gene signature possibly played a pivotal role in the status of the PAAD immune microenvironment.ConclusionA novel prognostic model based on immune genes may serve to characterize the immune microenvironment and provide a basis for PAAD immunotherapy.
| S-EPMC8318842 | biostudies-literature
Project description:PURPOSE:Stomach adenocarcinoma (STAD) is one of the most common malignant tumors, and its occurrence and prognosis are closely related to inflammation. The aim of the present study was to identify gene signatures and construct an immune-related gene (IRG) prognostic model in STAD using bioinformatics analysis. METHODS:RNA sequencing data from healthy samples and samples with STAD, IRGs, and transcription factors were analyzed. The hub IRGs were identified using univariate and multivariate Cox regression analyses. Using the hub IRGs, we constructed an IRG prognostic model. The relationships between IRG prognostic models and clinical data were tested. RESULTS:A total of 289 differentially expressed IRGs and 20 prognostic IRGs were screened with a threshold of P<0.05. Through multivariate stepwise Cox regression analysis, we developed a prognostic model based on seven IRGs. The prognostic model was validated using a GEO dataset (GSE 84437). The IRGs were significantly correlated with the clinical outcomes (age, histological grade, N, and M stage) of STAD patients. The infiltration abundances of dendritic cells and macrophages were higher in the high-risk group than in the low-risk group. CONCLUSIONS:Our results provide novel insights into the pathogenesis of STAD. An IRG prognostic model based on seven IRGs exhibited the predictive value, and have potential application value in clinical decision-making and individualized treatment.
| S-EPMC7593539 | biostudies-literature
Project description:BackgroundOvarian cancer is one of the most lethal diseases of women. The prognosis of ovarian cancer patients was closely correlated with immune cell expression and immune responses. Therefore, it is important to identify a robust prognostic signature, which correlates not only with prognoses but also with immune responses in ovarian cancer, thus, providing immune-related patient therapies.MethodsThe weighted gene coexpression network analysis (WGCNA) was used to identify candidate genes correlated with ovarian cancer prognoses. Univariate and multivariate Cox regression analyses were used to construct the prognostic signature. The Kaplan-Meier method was used to predict survival, and the immune-related bioinformatics analysis was performed using the R software. The relationship between the signature and clinical parameters was analyzed with the GraphPad Prism 7 and SPSS software.ResultsGene expression from The Cancer Genome Atlas dataset was used to perform the WGCNA analysis, and candidate prognostic-related genes in patients with ovarian cancer were identified. According to the Cox regression analysis, the prognostic signature was constructed, which divided patients into two groups. The high-risk group showed the least favorable prognosis. Three independent cohorts from the Gene Expression Omnibus (GEO) database were used for the validation studies. According to the immune analyses, the GEO database signatures were significantly correlated with the immune statuses of ovarian cancer patients. By analyzing the combination of the prognostic signature and total mutational burden (TMB), ovarian cancer patients were divided into four groups. In these groups, memory B cell, resting memory CD4 T cell, M2 macrophage, resting mast cell, and neutrophil were found with significant distinctions among these groups.ConclusionsThis novel signature predicted the prognosis of ovarian cancer patients precisely and independently and showed significant correlations with immune responses. Therefore, this prognostic signature could indicate targeted immunotherapies for ovarian cancer patients.
| S-EPMC7749778 | biostudies-literature
Project description:Pancreatic ductal adenocarcinoma (PDAC) is among the most lethal cancers. Known risk factors for this disease are currently insufficient in predicting mortality. In order to better prognosticate patients with PDAC, we identified 20 genes by utilizing publically available high-throughput transcriptomic data from GEO, TCGA and ICGC which are associated with overall survival and event-free survival. A score generated based on the expression matrix of these genes was validated in two independent cohorts. We find that this "Pancreatic cancer prognostic score 20 -PPS20" is independent of the confounding factors in multivariate analyses, is dramatically elevated in metastatic tissue compared to primary tumor, and is higher in primary tumors compared to normal pancreatic tissue. Transcriptomic analyses show that tumors with low PPS20 have overall more immune cell infiltration and a higher CD8 T cell/Treg ratio when compared to those with high PPS20. Analyses of proteomic data from TCGA PAAD indicated higher levels of Cyclin B1, RAD51, EGFR and a lower E-cadherin/Fibronectin ratio in tumors with high PPS20. The PPS20 score defines not only prognostic and biological sub-groups but can predict response to targeted therapy as well. Overall, PPS20 is a stronger and more robust transcriptomic signature when compared to similar, previously published gene lists.
| S-EPMC7170253 | biostudies-literature
Project description:BackgroundLung adenocarcinoma (LUAD) results in a majority of cancer burden worldwide. TP53 is the most commonly mutated in LUAD. This study aimed to reveal the relation between TP53 and tumor microenvironment (TME) for improving LUAD treatment.MethodsDifferentially expressed genes (DEGs) related to immunity were analyzed between TP53-WT and TP53-MUT groups. Least absolute shrinkage and selection operator (LASSO) Cox regression was applied to screen prognostic DEGs. Two independent datasets were included to evaluate the robustness of the prognostic model.ResultsAn 8-gene prognostic model containing ANLN, CCNB1, DLGAP5, FAM83A, GJB2, NAPSA, SFTPB, and SLC2A1 was established based on DEGs. LUAD samples were classified into high- and low-risk groups with differential overall survival in the two datasets. M0 macrophages, M1 macrophages, and activated memory CD4 T cells were more enriched in high-risk group. Immune checkpoints of PDCD1, LAG3, and CD274 were also high-expressed in high-risk group.ConclusionThe study improved the understanding of the role of TP53 in the TME modulation. The 8-gene model had robust performance to predict LUAD prognosis in clinical practice. In addition, the eight prognostic genes may also serve as potential targets for designing therapeutic drugs for LUAD patients.
| S-EPMC9279974 | biostudies-literature
Project description:The aim of current study was to use Weighted Gene Coexpression Network Analysis (WGCNA) to identify hub genes related to the incidence and prognosis of KRAS mutant (MT) lung adenocarcinoma (LUAD).We involved 184 stage IIB to IV LUAD samples and 59 normal lung tissue samples from The Cancer Genome Atlas (TCGA) database. The R package "limma" was used to identify differentially expressed genes (DEGs). WGCNA and survival analyses were performed by R packages "WGCNA" and "survival," respectively. The functional analyses were performed by R package "clusterProfiler" and GSEA software. Network construction and MCODE analysis were performed by Cytoscape_v3.6.1.Totally 2590 KRAS MT specific DEGs were found between LUAD and normal lung tissues, and 10 WGCNA modules were identified. Functional analysis of the key module showed the ribosome biogenesis related terms were enriched. We observed the expression of 8 genes were positively correlated to the worse survival of KRAS MT LUAD patients, the 7 of them were validated by Kaplan-Meier plotter database (kmplot.com/) (thymosin Beta 10 [TMSB10], ribosomal Protein S16 [RPS16], mitochondrial ribosomal protein L27 [MRPL27], cytochrome c oxidase subunit 6A1 [COX6A1], HCLS1-associated protein X-1 [HAX1], ribosomal protein L38 [RPL38], and ATP Synthase Membrane Subunit DAPIT [ATP5MD]). The GSEA analysis found mTOR and STK33 pathways were upregulated in KRAS MT LUAD (P < .05, false discovery rate [FDR] < 0.25).In summary, our study firstly used WGCNA to identify hub genes in the development of KRAS MT LUAD. The identified prognostic factors would be potential biomarkers in clinical use. Further molecular studies are required to confirm the mechanism of those genes in KRAS MT LUAD.
| S-EPMC7593058 | biostudies-literature
Project description:BackgroundPancreatic cancer has a 5-year overall survival lower than 8%. Pancreatic adenocarcinoma (PAAD) is the most common type. This study attempted to explore novel molecular subtypes and a prognostic model through analyzing tumor microenvironment (TME).Materials and methodsSingle-cell RNA sequencing (scRNA-seq) data and expression profiles from public databases were downloaded. Three PAAD samples with single-cell data and 566 samples with gene expression data were included. Seurat was used to identify cell subsets. SVA merged and removed batch effects from multichip datasets. CIBERSORT was used to evaluate the components of different cells in transcriptome, ConsensusClusterPlus was used to identify molecular subtypes, and gene set enrichment analysis was used for functional enrichment analysis. LASSO Cox was performed to construct dimensionality reduction and prognosis model.ResultsMemory B cells (MBCs) were identified to be significantly with PAAD prognosis. Two immune subtypes (IS1 and IS2) with distinct overall survival were constructed. Forty-one DEGs were identified between IS1 and IS2. Four prognostic genes (ANLN, ARNTL2, SERPINB5, and DKK1) were screened to develop a prognostic model. The model was effective in classifying samples into high-risk and low-risk groups with distinct prognosis. Three subgroups of MBCs were identified, where MBC_0 and MBC_1 were differentially distributed between IS1 and IS2, high-risk and low-risk groups.ConclusionsMBCs were closely involved in PAAD progression, especially MBC_0 and MBC_1 subgroups. The four-gene prognostic model was predictive of overall survival and could guide immunotherapy for patients with PAAD.
| S-EPMC9253429 | biostudies-literature