Improving Machine Learning Prediction of ADHD Using Gene Set Polygenic Risk Scores and Risk Scores from Genetically Correlated Phenotypes

Background: Polygenic risk scores (PRSs), which sum the effects of SNPs throughout the genome to measure risk afforded by common genetic variants, have improved our ability to estimate disorder risk for Attention-Deficit/Hyperactivity Disorder (ADHD) but the accuracy of risk prediction is rarely investigated. Methods: With the goal of improving risk prediction, we performed gene set analysis of GWAS data to select gene sets associated with ADHD within a training subset. For each selected gene set, we generated gene set polygenic risk scores (gsPRSs), which sum the effects of SNPs for each selected gene set. We created gsPRS for ADHD and for phenotypes having a high genetic correlation with ADHD. These gsPRS were added to the standard PRS as input to machine learning models predicting ADHD. We used feature importance scores to select gsPRS for a final model and to generate a ranking of the most consistently predictive gsPRS. Results: For a test subset that had not been used for training or validation, a random forest (RF) model using PRSs from ADHD and genetically correlated phenotypes and an optimized group of 20 gsPRS had an area under the receiving operating characteristic curve (AUC) of 0.72 (95% CI: 0.70 to 0.74). This AUC was a statistically significant improvement over logistic regression models and RF models using only PRS from ADHD and genetically correlated phenotypes. Conclusions: Summing risk at the gene set level and incorporating genetic risk from disorders with high genetic correlations with ADHD improved the accuracy of predicting ADHD. Learning curves suggest that additional improvements would be expected with larger study sizes. Our study suggests that better accounting of genetic risk and the genetic context of allelic differences results in more predictive models.

Identification and Validation of HOTAIRM1 as a Novel Biomarker for Oral Squamous Cell Carcinoma

ORAL squamous cell carcinoma (OSCC) is a malignant tumor with the highest incidence among tumors involving the oral cavity maxillofacial region, and is notorious for its high recurrence and metastasis potential. Long non-coding RNAs (lncRNAs), which regulate the genesis and evolution of cancers, are potential prognostic biomarkers. This study identified HOTAIRM1 as a novel significantly upregulated lncRNA in OSCC, which is strongly associated with unfavorable prognosis of OSCC. Systematic bioinformatics analyses demonstrated that HOTAIRM1 was closely related to tumor stage, overall survival, genome instability, the tumor cell stemness, the tumor microenvironment, and immunocyte infiltration. Using biological function prediction methods, including Weighted gene co-expression network analysis (WGCNA), Gene set enrichment analysis (GSEA), and Gene set variation analysis (GSVA), HOTAIRM1 plays a pivotal role in OSCC cell proliferation, and is mainly involved in the regulation of the cell cycle. In vitro, cell loss-functional experiments confirmed that HOTAIRM1 knockdown significantly inhibited the proliferation of OSCC cells, and arrested the cell cycle in G1 phase. At the molecular level, PCNA and CyclinD1 were obviously reduced after HOTAIRM1 knockdown. The expression of p53 and p21 was upregulated while CDK4 and CDK6 expression was decreased by HOTAIRM1 knockdown. In vivo, knocking down HOTAIRM1 significantly inhibited tumor growth, including the tumor size, weight, volume, angiogenesis, and hardness, monitored by ultrasonic imaging and magnetic resonance imaging In summary, our study reports that HOTAIRM1 is closely associated with tumorigenesis of OSCC and promotes cell proliferation by regulating cell cycle. HOTAIRM1 could be a potential prognostic biomarker and a therapeutic target for OSCC.

Identification of Novel Prognostic Markers Associated With Laryngeal Squamous Cell Carcinoma Using Comprehensive Analysis

BackgroundLaryngeal squamous cell carcinoma (LSCC) is a leading malignant cancer of the head and neck. Patients with LSCC, in which the cancer has infiltrated and metastasized, have a poor prognosis. Therefore, there is an urgent need to identify more potential targets for drugs and biomarkers for early diagnosis.MethodsRNA sequence data from LSCC and patients’ clinical traits were obtained from the Gene Expression Omnibus (GEO) (GSE142083) and The Cancer Genome Atlas (TCGA) database. Differentially expressed gene (DEG) analysis and weighted gene co-expression network analysis (WGCNA) were performed to identify hub genes. Gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis, prognostic value analysis, receiver operating characteristic (ROC) curve analysis, gene mutation analysis, tumor-infiltrating immune cell abundance profile estimation, gene set variation analysis (GSVA), and gene set enrichment analysis (GSEA) were performed. Single-gene RNA sequencing data were obtained from the GSE150321 dataset. Cell proliferation and viability were confirmed by the CCK-8 assay and real-time PCR.ResultsA total of 701 DEGs, including 329 upregulated and 372 downregulated genes, were screened in the GSE142083 dataset. Using WGCNA, three modules were identified to be closely related to LSCC. After intersecting the DEGs and performing univariate and multivariate Cox analyses, a novel prognostic model based on three genes (SLC35C1, HOXB7, and TEDC2) for LSCC was established. Interfering TEDC2 expression inhibited tumor cell proliferation and migration.ConclusionsOur results show that SLC35C1, HOXB7, and TEDC2 have the potential to become new therapeutic targets and prognostic biomarkers for LSCC.

Comprehensive exploration of the genetic contribution of the dopaminergic and serotonergic pathways to psychiatric disorders

Psychiatric disorders are highly prevalent and display considerable clinical and genetic overlap. Dopaminergic and serotonergic neurotransmission have been shown to play an important role in many psychiatric disorders. Here we aim to assess the genetic contribution of these systems to eight psychiatric disorders (attention-deficit hyperactivity disorder (ADHD), anorexia nervosa (ANO), autism spectrum disorder (ASD), bipolar disorder (BIP), major depression (MD), obsessive-compulsive disorder (OCD), schizophrenia (SCZ) and Tourette’s syndrome (TS)) using publicly available GWAS analyses performed by the Psychiatric Genomics Consortium that include more than 160,000 cases and 275,000 controls. To do so, we elaborated four different gene sets: two ‘wide’ selections for dopamine (DA) and for serotonin (SERT) using the Gene Ontology and KEGG pathways tools, and two’core’ selections for the same systems, manually curated. At the gene level, we found 67 genes from the DA and/or SERT gene sets significantly associated with one of the studied disorders, and 12 of them were associated with two different disorders. Gene-set analysis revealed significant associations for ADHD and ASD with the wide DA gene set, for BIP with the wide SERT gene set, and for MD with the core SERT set. Interestingly, interrogation of a cross-disorder GWAS meta-analysis of the eight psychiatric conditions displayed association with the wide DA gene set. To our knowledge, this is the first systematic examination of genes encoding proteins essential to the function of these two neurotransmitter systems in these disorders. Our results support a pleiotropic contribution of the dopaminergic and serotonergic systems in several psychiatric conditions.

Identification of Key Genes Associated with Endothelial Cell Dysfunction in Atherosclerosis Using Multiple Bioinformatics Tools

Atherosclerosis is the most notable cardiovascular disease, the latter being the main cause of death globally. Endothelial cell dysfunction plays a major role in the pathogenesis of atherosclerosis. However, it is currently unclear which genes are involved between endothelial cell dysfunction and atherosclerosis. This study was aimed at identifying these genes. Based on the GSE83500 dataset, the quantification of endothelial cell function was conducted using single-sample gene set enrichment analysis; the coexpression modules were conducted using weighted correlation network analysis. After building module-trait relationships, tan and yellow modules were regarded as hub modules. 10 hub genes from each hub module were identified by the protein-protein interaction network analysis. The key genes (RAB5A, CTTN, ITGB1, and MMP9) were obtained by comparing the expression differences of the hub gene between atherosclerotic and normal groups from the GSE28829 and GSE43292 datasets, respectively. ROC analysis showed the diagnostic value of key genes. Moreover, the differential expression of key genes in normal and atherosclerotic aortic walls was verified. In vitro, we establish a model of ox-LDL-injured endothelial cells and transfect RAB5A overexpression and shRNA plasmids. The results showed that overexpression of RAB5A ameliorates the proliferation and migration function of ox-LDL-injured endothelial cells, including the ability of tubule formation. It was speculated that the interferon response, Notch signaling pathways, etc. were involved in this function of RAB5A by using gene set variation analysis. With the multiple bioinformatics analysis methods, we detected that yellow and tan modules are related to the abnormal proliferation and migration of endothelial cells associated with atherosclerosis. RAB5A, CTTN, ITGB1, and MMP9 can be used as potential targets for therapy and diagnostic markers. In vitro, overexpression of RAB5A can ameliorate the proliferation and migration function of ox-LDL-injured endothelial cells, and the possible molecules involved in this process were speculated.

Comprehensive analysis of a TNF family based-signature in diffuse gliomas with regard to prognosis and immune significance

Background Several studies have shown that members of the tumor necrosis factor (TNF) family play an important role in cancer immunoregulation, and trials targeting these molecules are already underway. Our study aimed to integrate and analyze the expression patterns and clinical significance of TNF family-related genes in gliomas. Methods A total of 1749 gliomas from 4 datasets were enrolled in our study, including the Cancer Genome Atlas (TCGA) dataset as the training cohort and the other three datasets (CGGA, GSE16011, and Rembrandt) as validation cohorts. Clinical information, RNA expression data, and genomic profile were collected for analysis. We screened the signature gene set by Cox proportional hazards modelling. We evaluated the prognostic value of the signature by Kaplan–Meier analysis and timeROC curve. Gene Ontology (GO) and Gene set enrichment analysis (GSEA) analysis were performed for functional annotation. CIBERSORT algorithm and inflammatory metagenes were used to reveal immune characteristics. Results In gliomas, the expression of most TNF family members was positively correlated. Univariate analysis showed that most TNF family members were related to the overall survival of patients. Then through the LASSO regression model, we developed a TNF family-based signature, which was related to clinical, molecular, and genetic characteristics of patients with glioma. Moreover, the signature was found to be an independent prognostic marker through survival curve analysis and Cox regression analysis. Furthermore, a nomogram prognostic model was constructed to predict individual survival rates at 1, 3 and 5 years. Functional annotation analysis revealed that the immune and inflammatory response pathways were enriched in the high-risk group. Immunological analysis showed the immunosuppressive status in the high-risk group. Conclusions We developed a TNF family-based signature to predict the prognosis of patients with glioma.

Transcriptomics and network analysis highlight potential pathways in the pathogenesis of pterygium

Pterygium is a common ocular surface condition frequently associated with irritative symptoms. The precise identity of its critical triggers as well as the hierarchical relationship between all the elements involved in the pathogenesis of this disease are not yet elucidated. Meta-analysis of gene expression studies represents a novel strategy capable of identifying key pathogenic mediators and therapeutic targets in complex diseases. Samples from nine patients were collected during surgery after photo documentation and clinical characterization of pterygia. Gene expression experiments were performed using Human Clariom D Assay gene chip. Differential gene expression analysis between active and atrophic pterygia was performed using limma package after adjusting variables by age. In addition, a meta-analysis was performed including recent gene expression studies available at the Gene Expression Omnibus public repository. Two databases including samples from adults with pterygium and controls fulfilled our inclusion criteria. Meta-analysis was performed using the Rank Production algorithm of the RankProd package. Gene set analysis was performed using ClueGO and the transcription factor regulatory network prediction was performed using appropriate bioinformatics tools. Finally, miRNA-mRNA regulatory network was reconstructed using up-regulated genes identified in the gene set analysis from the meta-analysis and their interacting miRNAs from the Brazilian cohort expression data. The meta-analysis identified 154 up-regulated and 58 down-regulated genes. A gene set analysis with the top up-regulated genes evidenced an overrepresentation of pathways associated with remodeling of extracellular matrix. Other pathways represented in the network included formation of cornified envelopes and unsaturated fatty acid metabolic processes. The miRNA-mRNA target prediction network, also reconstructed based on the set of up-regulated genes presented in the gene ontology and biological pathways network, showed that 17 target genes were negatively correlated with their interacting miRNAs from the Brazilian cohort expression data. Once again, the main identified cluster involved extracellular matrix remodeling mechanisms, while the second cluster involved formation of cornified envelope, establishment of skin barrier and unsaturated fatty acid metabolic process. Differential expression comparing active pterygium with atrophic pterygium using data generated from the Brazilian cohort identified differentially expressed genes between the two forms of presentation of this condition. Our results reveal differentially expressed genes not only in pterygium, but also in active pterygium when compared to the atrophic ones. New insights in relation to pterygium’s pathophysiology are suggested.

ceRNA Network and Functional Enrichment Analysis of Preeclampsia by Weighted Gene Coexpression Network Analysis

Background. Preeclampsia (PE) is a multisystemic syndrome which has short- and long-term risk to mothers and children and has pluralistic etiology. Objective. This study is aimed at constructing a competitive endogenous RNA (ceRNA) network for pathways most related to PE using a data mining strategy based on weighted gene coexpression network analysis (WGCNA). Methods. We focused on pathways involving hypoxia, angiogenesis, and epithelial mesenchymal transition according to the gene set variation analysis (GSVA) scores. The gene sets of these three pathways were enriched by gene set enrichment analysis (GSEA). WGCNA was used to study the underlying molecular mechanisms of the three pathways in the pathogenesis of PE by analyzing the relationship among pathways and genes. The soft threshold power (β) and topological overlap matrix allowed us to obtain 15 modules, among which the red module was chosen for the downstream analysis. We chose 10 hub genes that satisfied ∣ log 2 Fold Change ∣ > 2 and had a higher degree of connectivity within the module. These candidate genes were subsequently confirmed to have higher gene significance and module membership in the red module. Coexpression networks were established for the hub genes to unfold the connection between the genes in the red module and PE. Finally, ceRNA networks were constructed to further clarify the underlying molecular mechanism involved in the occurrence of PE. 56 circRNAs, 17 lncRNAs, and 20 miRNAs participated in the regulation of the hub genes. Coagulation factor II thrombin receptor (F2R) and lumican (LUM) were considered the most relevant genes, and ceRNA networks of them were constructed. Conclusion. The microarray data mining process based on bioinformatics methods constructed lncRNA and miRNA networks for ten hub genes that were closely related to PE and focused on ceRNAs of F2R and LUM finally. The results of our study may provide insight into the mechanisms underlying PE occurrence.

ZCCHC17 Served as a Predictive Biomarker for Prognosis and Immunotherapy in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is one of the common malignant tumors. The prognosis and five-year survival rate of HCC are not promising due to tumor recurrence and metastasis. Exploring markers that contribute to the early diagnosis of HCC, markers for prognostic evaluation of HCC patients, and effective targets for treating HCC patients are in the spotlight of HCC therapy. Zinc Finger CCHC-Type Containing 17 (ZCCHC17) encodes the RNA binding protein ZCCHC17, but its role in HCC is still unclear. Here, 90 paraffin-embedded specimens combined with bioinformatics were used to comprehensively clarify the value of ZCCHC17 in the diagnosis and prognosis of HCC and its potential functions. Paraffin-embedded specimens were used to assess ZCCHC17 protein expression and its correlation with prognosis in 90 HCC patients. the public data sets of HCC patients from TCGA, ICG, and GEO databases were also used for further analysis. It was found that protein and mRNA levels of ZCCHC17 in HCC tissues were significantly higher than those in normal tissues. The abnormally high expression may be related to the abnormal DNA methylation of ZCCHC17 in tumor tissues. The high expression of ZCCHC17 is related to AFP, histologic grade, tumor status, vascular invasion, and pathological stage. Multi-data set analysis showed that patients with high ZCCHC17 expression had a worse prognosis, and multivariate cox regression analysis showed an independent prognostic significance of ZCCHC17. The results of functional analysis, including Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Set Enrichment Analysis (GSEA), indicate that ZCCHC17 is mainly involved in immune regulation. Subsequently, further single-sample gene set enrichment analysis (ssGSEA) showed that the expression of ZCCHC17 was related to the infiltration of immune cells. Importantly, we also analyzed the relationship between ZCCHC17 and immune checkpoint genes, tumor mutation burden (TMB), microsatellite instability (MSI) and TP53 status in HCC patients and evaluated the role of ZCCHC17 in cancer immunotherapy. In summary, ZCCHC17 is a novel marker for the diagnosis and prognostic evaluation of HCC. Concurrently, it regulates immune cells in the tumor microenvironment (TME) of HCC patients, which has a specific reference value for the immunotherapy of HCC.