Identification of Candidate Forage Yield Genes in Sorghum (Sorghum bicolor L.) Using Integrated Genome-Wide Association Studies and RNA-Seq

Genetic dissection of forage yield traits is critical to the development of sorghum as a forage crop. In the present study, association mapping was performed with 85,585 SNP markers on four forage yield traits, namely plant height (PH), tiller number (TN), stem diameter (SD), and fresh weight per plant (FW) among 245 sorghum accessions evaluated in four environments. A total of 338 SNPs or quantitative trait nucleotides (QTNs) were associated with the four traits, and 21 of these QTNs were detected in at least two environments, including four QTNs for PH, ten for TN, six for SD, and one for FW. To identify candidate genes, dynamic transcriptome expression profiling was performed at four stages of sorghum development. One hundred and six differentially expressed genes (DEGs) that were enriched in hormone signal transduction pathways were found in all stages. Weighted gene correlation network analysis for PH and SD indicated that eight modules were significantly correlated with PH and that three modules were significantly correlated with SD. The blue module had the highest positive correlation with PH and SD, and the turquoise module had the highest negative correlation with PH and SD. Eight candidate genes were identified through the integration of genome-wide association studies (GWAS) and RNA sequencing. Sobic.004G143900, an indole-3-glycerol phosphate synthase gene that is involved in indoleacetic acid biosynthesis, was down-regulated as sorghum plants grew in height and was identified in the blue module, and Sobic.003G375100, an SD candidate gene, encoded a DNA repair RAD52-like protein 1 that plays a critical role in DNA repair-linked cell cycle progression. These findings demonstrate that the integrative analysis of omics data is a promising approach to identify candidate genes for complex traits.

Identification of Cross-Talk and Pyroptosis-Related Genes Linking Periodontitis and Rheumatoid Arthritis Revealed by Transcriptomic Analysis

Background. The current study is aimed at identifying the cross-talk genes between periodontitis (PD) and rheumatoid arthritis (RA), as well as the potential relationship between cross-talk genes and pyroptosis-related genes. Methods. Datasets for the PD (GSE106090, GSE10334, GSE16134) and RA (GSE55235, GSE55457, GSE77298, and GSE1919) were downloaded from the GEO database. After batch correction and normalization of datasets, differential expression analysis was performed to identify the differentially expressed genes (DEGs). The cross-talk genes linking PD and RA were obtained by overlapping the DEGs dysregulated in PD and DEGs dysregulated in RA. Genes involved in pyroptosis were summarized by reviewing literatures, and the correlation between pyroptosis genes and cross-talk genes was investigated by Pearson correlation coefficient. Furthermore, the weighted gene coexpression network analysis (WGCNA) was carried out to identify the significant modules which contained both cross-talk genes and pyroptosis genes in both PD data and RA data. Thus, the core cross-talk genes were identified from the significant modules. Receiver-operating characteristic (ROC) curve analysis was performed to identify the predictive accuracy of these core cross-talk genes in diagnosing PD and RA. Based on the core cross-talk genes, the experimentally validated protein-protein interaction (PPI) and gene-pathway network were constructed. Results. A total of 40 cross-talk genes were obtained. Most of the pyroptosis genes were not differentially expressed in disease and normal samples. By selecting the modules containing both cross-talk genes or pyroptosis genes, the blue module was identified to be significant module. Three genes, i.e., cross-talk genes (TIMP1, LGALS1) and pyroptosis gene-GPX4, existed in the blue module of PD network, while two genes (i.e., cross-talk gene-VOPP1 and pyroptosis gene-AIM2) existed in the blue module of RA network. ROC curve analysis showed that three genes (TIMP1, VOPP1, and AIM2) had better predictive accuracy in diagnosing disease compared with the other two genes (LGALS1 and GPX4). Conclusions. This study revealed shared mechanisms between RA and PD based on cross-talk and pyroptosis genes, supporting the relationship between the two diseases. Thereby, five modular genes (TIMP1, LGALS1, GPX4, VOPP1, and AIM2) could be of relevance and might serve as potential biomarkers. These findings are a basis for future research in the field.

Metabolic syndrome related gene signature predicts the prognosis of patients with pancreatic ductal carcinoma. A novel link between metabolic dysregulation and pancreatic ductal carcinoma

Background Pancreatic cancer is one of the most common malignancies worldwide. In recent years, specific metabolic activities, which involves the development of tumor, caused wide public concern. In this study, we wish to explore the correlation between metabolism and progression of tumor. Methods A retrospective analysis including 95 patients with pancreatic ductal adenocarcinoma (PDAC) and PDAC patients from The Cancer Genome Atlas (TCGA), the International Cancer Genome Consortium (ICGC), and The Gene Expression Omnibus (GEO) database were involved in our study. Multivariate Cox regression analysis was used to construct the prognosis model. The potential connection between metabolism and immunity of PDAC was investigated through a weighted gene co-expression network analysis (WGCNA). 22 types of Tumor-infiltrating immune cells (TIICs) between high-risk and low-risk groups were estimated through CIBERSORT. Moreover, the potential immune-related signaling pathways between high-risk and low-risk groups were explored through the gene set enrichment analysis (GSEA). The role of key gene GMPS in developing pancreatic tumor was further investigated through CCK-8, colony-information, and Transwell. Results The prognostic value of the MetS factors was analyzed using the Cox regression model, and a clinical MetS-based nomogram was established. Then, we established a metabolism-related signature to predict the prognosis of PDAC patients based on the TCGA databases and was validated in the ICGC database and the GEO database to find the distinct molecular mechanism of MetS genes in PDAC. The result of WGCNA showed that the blue module was associated with risk score, and genes in the blue module were found to be enriched in the immune-related signaling pathway. Furthermore, the result of CIBERSORT demonstrated that proportions of T cells CD8, T cells Regulatory, Tregs NK cells Activated, Dendritic cells Activated, and Mast cells Resting were different between high-risk and low-risk groups. These differences are potential causes of different prognoses of PDAC patients. GSEA and the protein–protein interaction network (PPI) further revealed that our metabolism-related signature was significantly enriched in immune‐related biological processes. Moreover, knockdown of GMPS in PDAC cells suppressed proliferation, migration, and invasion of tumor cells, whereas overexpression of GMPS performed oppositely. Conclusion The results shine light on fundamental mechanisms of metabolic genes on PDAC and establish a reliable and referable signature to evaluate the prognosis of PDAC. GMPS was identified as a potential candidate oncogene with in PDAC, which can be a novel biomarker and therapeutic target for PDAC treatment.

FAM201A, a Long Noncoding RNA Potentially Associated With Atrial Fibrillation Identified by ceRNA Network Analyses and WGCNA

Background: Atrial fibrillation (AF) is the most common cardiac arrhythmia that contributes to various complications. However, little is known about lncRNAs associated with AF susceptibility. In the present study, we aim to identify lncRNAs involved in pathogenesis of AF based on competing endogenous RNA (ceRNA) network analyses and weighted gene co-expression network analysis (WGCNA).Methods: Two lncRNA and mRNA microarray datasets GSE41177 and GSE79768 were retrieved from the Gene Expression Omnibus (GEO) database. Differentially expressed lncRNAs (DElncRNAs), mRNAs (DEmRNAs) between AF patients and patients with sinus rhythm (SR) were identified from dataset GSE41177. Then, those DElncRNAs associated target miRNAs were predicted. The ceRNA network was constructed based on DElncRNAs, predicted miRNAs and DEmRNAs. To validate the role of AF-related lncRNAs, all lncRNAs form dataset GSE79768 were selected to perform WGCNA. LncRNA modules relevant to AF were identified. Crucial lncRNAs in the module that was most relevant to AF were screened according to the criteria of | Gene significance (GS)| > 0.6 and |Module membership (MM)| > 0.5. Results: A total of 18 DElncRNAs and 350 DEmRNAs were identified between AF patients and SR patients. The final ceRNA network contained 5 lncRNAs, 10 miRNAs, and 21 mRNAs. According to the ceRNA theory, combined with the comparative toxicogenomics database (CTD) database, the ceRNA axis FAM201A-miR-33a-3p-RAC3 was considered associated with AF susceptibility. By WGCNA, the blue module was detected most highly relevant with AF. The lncRNA FAM201A was proved in the blue module and highly related to AF. Conclusions: These results demonstrated that FAM201A might have great potential for susceptibility of AF based on ceRNA network analyses and WGCNA. FAM201A may function, at least partly, as ceRNA to regulate RAC3 in AF susceptibility.

Development and Validation of a Novel Ferroptosis-Related Gene Signature for Predicting Prognosis and the Immune Microenvironment in Gastric Cancer

Ferroptosis is a mode of regulated cell death that depends on iron and plays pivotal roles in regulating various biological processes in human cancers. However, the role of ferroptosis in gastric cancer (GC) remains unclear. In our study, a total of 2721 differentially expressed genes (DEGs) were filtered based on The Cancer Genome Atlas (TCGA) ( n = 375 ) dataset. Weighted gene coexpression network (WGCNA) analysis was then used and identified 7 modules, of which the blue module with the most significant enrichment result was selected. By taking the intersections of the blue module and ferroptosis-related genes (FRGs), we obtained 23 common genes. Functional analysis was performed to explore the biological function of the genes of interest, and with univariate Cox regression (UCR) analysis, survival genes were screened to construct a prognostic model based on 3 genes (SLC1A5, ANGPTL4, and CGAS), which could play a role in predicting the survival of GC patients. UCR and multivariate Cox regression (MCR) analysis revealed that the prognostic index could be used as an independent prognostic indicator and validated using another GSE84437 dataset. Notably, patients in the high-risk group had higher mutation frequencies, such as TTN and TP53. TIMER analysis demonstrated that the risk score strongly correlated with macrophage and CD4+ T cell infiltration. In addition, the high- and low-risk groups illustrated different distributions of different immune statuses. Furthermore, the low-risk group had a higher immunophenoscore (IPS), which meant a better response to immune checkpoint inhibitors (ICIs). Finally, gene set enrichment analysis (GSEA) revealed several significant pathways involved in GC. In this study, a novel FRG signature was built that could predict GC prognosis and reflect the status of the tumor immune microenvironment.

Immune Mechanism, Gene Module, and Molecular Subtype Identification of Astragalus Membranaceus in the Treatment of Dilated Cardiomyopathy: An Integrated Bioinformatics Study

Astragalus membranaceus has complex components as a natural drug and has multilevel, multitarget, and multichannel effects on dilated cardiomyopathy (DCM). However, the immune mechanism, gene module, and molecular subtype of astragalus membranaceus in the treatment of DCM are still not revealed. Microarray information of GSE84796 was downloaded from the GEO database, including RNA sequencing data of seven normal cardiac tissues and ten DCM cardiac tissues. A total of 4029 DCM differentially expressed genes were obtained, including 1855 upregulated genes and 2174 downregulated genes. GO/KEGG/GSEA analysis suggested that the activation of T cells and B cells was the primary cause of DCM. WGCNA was used to obtain blue module genes. The blue module genes are primarily ADCY7, BANK1, CD1E, CD19, CD38, CD300LF, CLEC4E, FLT3, GPR18, HCAR3, IRF4, LAMP3, MRC1, SYK, and TLR8, which successfully divided DCM into three molecular subtypes. Based on the CIBERSORT algorithm, the immune infiltration profile of DCM was analyzed. Many immune cell subtypes, including the abovementioned immune cells, showed different levels of increased infiltration in the myocardial tissue of DCM. However, this infiltration pattern was not obviously correlated with clinical characteristics, such as age, EF, and sex. Based on network pharmacology and ClueGO, 20 active components of Astragalus membranaceus and 40 components of DMCTGS were obtained from TCMSP. Through analysis of the immune regulatory network, we found that Astragalus membranaceus effectively regulates the activation of immune cells, such as B cells and T cells, cytokine secretion, and other processes and can intervene in DCM at multiple components, targets, and levels. The above mechanisms were verified by molecular docking results, which confirmed that AKT1, VEGFA, MMP9, and RELA are promising potential targets of DCM.

Molecular Characteristics of Choledochal Cysts in Children: Transcriptome Sequencing

A choledochal cyst (CC) is a common congenital biliary disease in children, yet the underlying molecular bases for the cystic and fusiform clinical subtypes are unknown. RNA sequencing (RNA-seq) has been performed on 22 high-quality CC samples, including 12 cystic CC and 10 fusiform CC samples, to search for molecular features. Weighted gene co-expression network analysis (WGCNA) was performed to identify key modules associated with clinical subtypes. Bioinformatic analyses were conducted to elucidate potential mechanisms. Then, we constructed protein–protein interaction (PPI) networks to identify candidate hub genes related to CC. Finally, we used the support vector machine (SVM) to eliminate redundant features and screen out the hub genes. The selected gene expression was determined in CC patients through quantitative real-time polymerase chain reaction (PCR). A total of 6,463 genes were found to be aberrantly expressed between cystic CC and fusiform CC. Twelve co-expression modules that correlated with clinical subtypes of CC were identified and assigned representative colors. Among the 12 modules, the blue module was considered the key module. Two functionally distinct sets of dysregulated genes have been identified in two major subtypes, metabolism-related genes in cystic CC and immune-related genes in fusiform CC. A total of 20 candidate hub genes that were correlated with clinical subtypes were found in the blue module. In addition, we found ERBB2 and WNT11 that have not been studied in CC and verified their differential expression in CC through quantitative real-time PCR experiments. For the first time, we have described the transcriptome characteristics of CC. These results suggest that cystic CC and fusiform CC have different molecular mechanisms. The bi-omics-identified novel candidate genes and pathways might be helpful for personalized treatment and are of great clinical significance for CC.

Metabolic Syndrome Related Gene Signature Predicts the Prognosis of Patients With Pancreatic Ductal Carcinoma. A Novel Link Between Metabolic Dysregulation and Pancreatic Ductal Carcinoma.

Background: Pancreatic cancer is one of the most common malignancies worldwide. In recent years, specific metabolic activities, which involves the development of tumor, caused wide public concern. In this study, we wish to explore the correlation between metabolism and progression of tumor.Methods: A retrospective analysis including 95 patients with pancreatic ductal adenocarcinoma (PDAC) and PDAC patients from TCGA, ICGA, and GEO database were involved in our study. Multivariate Cox regression analysis was used to construct the prognosis model. The potential connection between metabolism and immunity of PDAC was investigated through WGCNA. 22 types of TIICs between high-risk and low-risk groups were estimated through CIBERSORT. Moreover, the potential immune-related signaling pathways between high-risk and low-risk groups were explored through the GSEA. The role of key gene GMPS in developing pancreatic tumor was further investigated through CCK-8, colony-information, and Transwell. Results: The prognostic value of the MetS factors was analyzed using the Cox regression model, and a clinical MetS-based nomogram was established. Then, we established a metabolism-related signature to predict the prognosis of PDAC patients based on the TCGA databases and was validated in the ICGC database and the GEO database to find the distinct molecular mechanism of MetS genes in PDAC. The result of WGCNA showed that the blue module was associated with risk score, and genes in the blue module were found to be enriched in the immune-related signaling pathway. Furthermore, the result of CIBERSORT demonstrated that proportions of T cells CD8, T cells Regulatory, Tregs NK cells Activated, Dendritic cells Activated, and Mast cells Resting were different between high-risk and low-risk groups. These differences are potential causes of different prognoses of PDAC patients. GSEA and the protein–protein interaction network (PPI) further revealed that our metabolism-related signature was significantly enriched in immune‐related biological processes. Moreover, knockdown of GMPS in PDAC cells suppressed proliferation, migration, and invasion of tumor cells, whereas overexpression of GMPS performed oppositely.Conclusion: The results shine light on fundamental mechanisms of metabolic genes on PDAC and establish a reliable and referable signature to evaluate the prognosis of PDAC. GMPS was identified as a potential candidate oncogene with in PDAC, which can be a novel biomarker and therapeutic target for PDAC treatment.

Identification the ferroptosis-related gene signature in Gastric cancer based on weighted gene co-expression network analysis (WGCNA)

Background Ferroptosis is a mode of regulated cell death that depends on iron, plays pivotal roles in regulating various biological process in human cancers. However, the role of ferroptosis in Gastric cancer (GC) remains unclear. Methods A total of 2721 differentially expressed genes (DEGs) were filtered base on The Cancer Genome Atlas (TCGA) (n = 375) dataset. Gene modules were identified based on co-expression network analysis (WGCNA). Functional analysis was performed to explore the biological function. Lasso-penalized and univariate Cox regression (UCR) analysis, survival genes were screened out to construct a prognostic model, which validated by the GSE43292 dataset. Gene set enrichment analysis (GSEA) for prognostic index was performed. Finally, the correlations of ferroptosis and immune cells were assessed through the TIMER database. Results Compared to normal specimens, 1063 highly upregulated and 1658 downregulated genes respectively and their normal counterparts in GC specimens were screened. WGCNA analysis was used and identified 7 modules, of which, blue module with the most significant enrichment result was selected. By taking intersections of blue module and differentially expressed ferroptosis-related genes (DEFRGs), we got 23 common genes. Functional analysis was performed to explore the biological function of the interested genes, and with the consequences Lasso-penalized and univariate Cox regression (UCR) analysis, survival genes were screened out to construct a prognostic model based on 3 genes (SLC1A5, ANGPTL4, and CGAS), which could play a role in predicting the survival of GC patients. UCR and multivariate Cox regression (MCR) analysis revealed that the prognostic index could be used as independent prognostic indicators and validated using another GSE84437 dataset. Notably, patients in high-risk groups had higher levels of higher mutation frequencies such as TTN and TP53.Mechanistically. Gene set enrichment analysis (GSEA) unveiled several significant and pathways involved in GC. TIMER analysis demonstrated that risk score strongly correlated with Macrophage and CD4 + T cells infiltration. In addition, high- and low-risk group illustrated different distributions in different immune status. Conclusions In this study, a novel FRGs signature was built. It could accurately predict GC prognosis and pave the new way for diagnosis and therapy strategy. May reflect the status of tumor immune microenvironment.

Identifying Conserved Generic Aspergillus spp. Co-Expressed Gene Modules Associated with Germination Using Cross-Platform and Cross-Species Transcriptomics

Aspergillus spp. is an opportunistic human pathogen that may cause a spectrum of pulmonary diseases. In order to establish infection, inhaled conidia must germinate, whereby they break dormancy, start to swell, and initiate a highly polarized growth process. To identify critical biological processes during germination, we performed a cross-platform, cross-species comparative analysis of germinating A. fumigatus and A. niger conidia using transcriptional data from published RNA-Seq and Affymetrix studies. A consensus co-expression network analysis identified four gene modules associated with stages of germination. These modules showed numerous shared biological processes between A. niger and A. fumigatus during conidial germination. Specifically, the turquoise module was enriched with secondary metabolism, the black module was highly enriched with protein synthesis, the darkgreen module was enriched with protein fate, and the blue module was highly enriched with polarized growth. More specifically, enriched functional categories identified in the blue module were vesicle formation, vesicular transport, tubulin dependent transport, actin-dependent transport, exocytosis, and endocytosis. Genes important for these biological processes showed similar expression patterns in A. fumigatus and A. niger, therefore, they could be potential antifungal targets. Through cross-platform, cross-species comparative analysis, we were able to identify biologically meaningful modules shared by A. fumigatus and A. niger, which underscores the potential of this approach.