irGSEA: the integration of single-cell rank-based gene set enrichment analysis.

irGSEA is an R package designed to assess the outcomes of various gene set scoring methods when applied to single-cell RNA sequencing data. This package incorporates six distinct scoring methods that rely on the expression ranks of genes, emphasizing relative expression levels over absolute values. The implemented methods include AUCell, UCell, singscore, ssGSEA, JASMINE and Viper. Previous studies have demonstrated the robustness of these methods to variations in dataset size and composition, generating enrichment scores based solely on the relative gene expression of individual cells. By employing the robust rank aggregation algorithm, irGSEA amalgamates results from all six methods to ascertain the statistical significance of target gene sets across diverse scoring methods. The package prioritizes user-friendliness, allowing direct input of expression matrices or seamless interaction with Seurat objects. Furthermore, it facilitates a comprehensive visualization of results. The irGSEA package and its accompanying documentation are accessible on GitHub (https://github.com/chuiqin/irGSEA).

GeneRaMeN enables integration, comparison, and meta-analysis of multiple ranked gene lists to identify consensus, unique, and correlated genes.

High-throughput experiments often produce ranked gene outputs, with forward genetic screening being a notable example. While there are various tools for analyzing individual datasets, those that perform comparative and meta-analytical examination of such ranked gene lists remain scarce. Here, we introduce Gene Rank Meta Analyzer (GeneRaMeN), an R Shiny tool utilizing rank statistics to facilitate the identification of consensus, unique, and correlated genes across multiple hit lists. We focused on two key topics to showcase GeneRaMeN: virus host factors and cancer dependencies. Using GeneRaMeN 'Rank Aggregation', we integrated 24 published and new flavivirus genetic screening datasets, including dengue, Japanese encephalitis, and Zika viruses. This meta-analysis yielded a consensus list of flavivirus host factors, elucidating the significant influence of cell line selection on screening outcomes. Similar analysis on 13 SARS-CoV-2 CRISPR screening datasets highlighted the pivotal role of meta-analysis in revealing redundant biological pathways exploited by the virus to enter human cells. Such redundancy was further underscored using GeneRaMeN's 'Rank Correlation', where a strong negative correlation was observed for host factors implicated in one entry pathway versus the alternate route. Utilizing GeneRaMeN's 'Rank Uniqueness', we analyzed human coronaviruses 229E, OC43, and SARS-CoV-2 datasets, identifying host factors uniquely associated with a defined subset of the screening datasets. Similar analyses were performed on over 1000 Cancer Dependency Map (DepMap) datasets spanning 19 human cancer types to reveal unique cancer vulnerabilities for each organ/tissue. GeneRaMeN, an efficient tool to integrate and maximize the usability of genetic screening datasets, is freely accessible via https://ysolab.shinyapps.io/GeneRaMeN.

Multiblock partial least squares and rank aggregation: Applications to detection of bacteriophages associated with antimicrobial resistance in the presence of potential confounding factors.

Urban environments, characterized by bustling mass transit systems and high population density, host a complex web of microorganisms that impact microbial interactions. These urban microbiomes, influenced by diverse demographics and constant human movement, are vital for understanding microbial dynamics. We explore urban metagenomics, utilizing an extensive dataset from the Metagenomics & Metadesign of Subways & Urban Biomes (MetaSUB) consortium, and investigate antimicrobial resistance (AMR) patterns. In this pioneering research, we delve into the role of bacteriophages, or "phages"-viruses that prey on bacteria and can facilitate the exchange of antibiotic resistance genes (ARGs) through mechanisms like horizontal gene transfer (HGT). Despite their potential significance, existing literature lacks a consensus on their significance in ARG dissemination. We argue that they are an important consideration. We uncover that environmental variables, such as those on climate, demographics, and landscape, can obscure phage-resistome relationships. We adjust for these potential confounders and clarify these relationships across specific and overall antibiotic classes with precision, identifying several key phages. Leveraging machine learning tools and validating findings through clinical literature, we uncover novel associations, adding valuable insights to our comprehension of AMR development.

A network biology approach to identify crucial host targets for COVID-19.

The ongoing global pandemic of COVID-19, caused by SARS-CoV-2 has killed more than 5.9 million individuals out of ∼43 million confirmed infections. At present, several parts of the world are encountering the 3rd wave. Mass vaccination has been started in several countries but they are less likely to be broadly available for the current pandemic, repurposing of the existing drugs has drawn highest attention for an immediate solution. A recent publication has mapped the physical interactions of SARS-CoV-2 and human proteins by affinity-purification mass spectrometry (AP-MS) and identified 332 high-confidence SARS-CoV-2-human protein-protein interactions (PPIs). Here, we taken a network biology approach and constructed a human protein-protein interaction network (PPIN) with the above SARS-CoV-2 targeted proteins. We utilized a combination of essential network centrality measures and functional properties of the human proteins to identify the critical human targets of SARS-CoV-2. Four human proteins, namely PRKACA, RHOA, CDK5RAP2, and CEP250 have emerged as the best therapeutic targets, of which PRKACA and CEP250 were also found by another group as potential candidates for drug targets in COVID-19. We further found candidate drugs/compounds, such as guanosine triphosphate, remdesivir, adenosine monophosphate, MgATP, and H-89 dihydrochloride that bind the target human proteins. The urgency to prevent the spread of infection and the death of diseased individuals has prompted the search for agents from the pool of approved drugs to repurpose them for COVID-19. Our results indicate that host targeting therapy with the repurposed drugs may be a useful strategy for the treatment of SARS-CoV-2 infection.

Identification of Signature Genes of Dilated Cardiomyopathy Using Integrated Bioinformatics Analysis.

Dilated cardiomyopathy (DCM) is characterized by left ventricular or biventricular enlargement with systolic dysfunction. To date, the underlying molecular mechanisms of dilated cardiomyopathy pathogenesis have not been fully elucidated, although some insights have been presented. In this study, we combined public database resources and a doxorubicin-induced DCM mouse model to explore the significant genes of DCM in full depth. We first retrieved six DCM-related microarray datasets from the GEO database using several keywords. Then we used the "LIMMA" (linear model for microarray data) R package to filter each microarray for differentially expressed genes (DEGs). Robust rank aggregation (RRA), an extremely robust rank aggregation method based on sequential statistics, was then used to integrate the results of the six microarray datasets to filter out the reliable differential genes. To further improve the reliability of our results, we established a doxorubicin-induced DCM model in C57BL/6N mice, using the "DESeq2" software package to identify DEGs in the sequencing data. We cross-validated the results of RRA analysis with those of animal experiments by taking intersections and identified three key differential genes (including BEX1, RGCC and VSIG4) associated with DCM as well as many important biological processes (extracellular matrix organisation, extracellular structural organisation, sulphur compound binding, and extracellular matrix structural components) and a signalling pathway (HIF-1 signalling pathway). In addition, we confirmed the significant effect of these three genes in DCM using binary logistic regression analysis. These findings will help us to better understand the pathogenesis of DCM and may be key targets for future clinical management.

A Comparative Study of Rank Aggregation Methods in Recommendation Systems.

The aim of a recommender system is to suggest to the user certain products or services that most likely will interest them. Within the context of personalized recommender systems, a number of algorithms have been suggested to generate a ranking of items tailored to individual user preferences. However, these algorithms do not generate identical recommendations, and for this reason it has been suggested in the literature that the results of these algorithms can be combined using aggregation techniques, hoping that this will translate into an improvement in the quality of the final recommendation. In order to see which of these techniques increase the quality of recommendations to the greatest extent, the authors of this publication conducted experiments in which they considered five recommendation algorithms and 20 aggregation methods. The research was carried out on the popular and publicly available MovieLens 100k and MovieLens 1M datasets, and the results were confirmed by statistical tests.

Quantiles based personalized treatment selection for multivariate outcomes and multiple treatments.

In this work, we propose a method for individualized treatment selection when there are correlated multiple responses for the K treatment ( K≥2 ) scenario. Here we use ranks of quantiles of outcome variables for each treatment conditional on patient-specific scores constructed from collected covariate measurements. Our method covers any number of treatments and outcome variables using any number of quantiles and it can be applied for a broad set of models. We propose a rank aggregation technique for combining several lists of ranks where both these lists and elements within each list can be correlated. The method has the flexibility to incorporate patient and clinician preferences into the optimal treatment decision on an individual case basis. A simulation study demonstrates the performance of the proposed method in finite samples. We also present illustrations using two different datasets from diabetes and HIV-1 clinical trials to show the applicability of the proposed procedure for real data.

Integrated analysis of transcriptomics to identify hub genes in primary Sjögren's syndrome.

OBJECTIVE: The treatment of patients with primary Sjögren's syndrome is a clinical challenge. Gene expression profile analysis and comprehensive network methods for complex diseases can provide insight into molecular characteristics in the clinical context. MATERIALS AND METHODS: We downloaded gene expression datasets from the Gene Expression Omnibus (GEO) database. We screened differentially expressed genes (DEG) between the pSS patients and the controls by the robust rank aggregation (RRA) method. We explored DEGs' potential function using gene function annotation and PPI network analysis. RESULTS: GSE23117, GSE40611, GSE80805, and GSE127952 were included, including 38 patients and 30 controls. The RRA integrated analysis determined 294 significant DEGs (241 upregulated and 53 downregulated), and the most significant gene aberrantly expressed in SS was CXCL9 (p = 6.39E-15), followed by CXCL13 (p = 1.53E-13). Immune response (GO:0006955; p = 4.29E-32) was the most significantly enriched biological process in GO (gene ontology) analysis. KEGG pathway enrichment analysis showed that cytokine-cytokine receptor interaction (hsa04060; p = 6.46E-10) and chemokine signaling pathway (hsa04062; p = 9.54E-09) were significantly enriched. We defined PTPRC, CD86, and LCP2 as the hub genes based on the PPI results. CONCLUSION: Our integrated analysis identified gene signatures and helped understand molecular changes in pSS.

Identification of biomarkers, immune infiltration landscape, and treatment targets of ischemia-reperfusion acute kidney injury at an early stage by bioinformatics methods.

BACKGROUND: Mechanisms underlying ischemia/reperfusion injury-acute kidney injury (IRI-AKI) are not fully elucidated. We conducted an integrative analysis of IRI-AKI by bioinformatics methods. METHODS: We screened gene expression profiles of the IRI-AKI at early phase from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) were identified and enrichment pathways were conducted based on gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) database, and Gene set enrichment analysis (GSEA). Immune cell infiltration analysis was performed to reveal the change of the microenvironment cell types. We constructed protein-protein interaction (PPI), and Cytoscape with plug-ins to find hub genes and modules. We performed robust rank aggregation (RRA) to combine DEGs and analyzed the target genes for miRNA/transcription factor (TF) and drug-gene interaction networks. RESULTS: A total of 239 and 384 DEGs were identified in GSE87024 and GSE34351 separately, with the 73 common DEGs. Enrichment analysis revealed that the significant pathways involve mitogen-activated protein kinase (MAPK) signaling, interleukin-17, and tumor necrosis factor (TNF) signaling pathway, etc. RRA analysis detected a total of 27 common DEGs. Immune cell infiltration analysis showed the plasma cells reduced and T cells increased in IRI-AKI. We identified JUN, ATF3, FOS, EGR1, HMOX1, DDIT3, JUNB, NFKBIZ, PPP1R15A, CXCL1, ATF4, and HSPA1B as hub genes. The target genes interacted with 23 miRNAs and 116 drugs or molecular compounds such as curcumin, staurosporine, and deferoxamine. CONCLUSION: Our study first focused on the early IRI-AKI adopting RRA analysis to combine DEGs in different datasets. We identified significant biomarkers and crucial pathways involved in IRI-AKI and first construct the immune landscape and detected the potential therapeutic targets of the IRI-AKI by drug-gene network.

Identification of four novel hub genes as monitoring biomarkers for colorectal cancer.

BACKGROUND: It must be admitted that the incidence of colorectal cancer (CRC) was on the rise all over the world, but the related treatment had not caught up. Further research on the underlying pathogenesis of CRC was conducive to improving the survival status of current CRC patients. METHODS: Differentially expressed genes (DEGs) screening were conducted based on "limma" and "RobustRankAggreg" package of R software. Weighted gene co-expression network analysis (WGCNA) was performed in the integrated DEGs that from The Cancer Genome Atlas (TCGA), and all samples of validation were from Gene Expression Omnlbus (GEO) dataset. RESULTS: The terms obtained in the functional annotation for primary DEGs indicated that they were associated with CRC. The MEyellow stand out whereby showed the significant correlation with clinical feature (disease), and 4 hub genes, including ABCC13, AMPD1, SCNN1B and TMIGD1, were identified in yellow module. Nine datasets from Gene Expression Omnibus database confirmed these four genes were significantly down-regulated and the survival estimates for the low-expression group of these genes were lower than for the high-expression group in Kaplan-Meier survival analysis section. MEXPRESS suggested that down-regulation of some top hub genes may be caused by hypermethylation. Receiver operating characteristic curves indicated that these genes had certain diagnostic efficacy. Moreover, tumor-infiltrating immune cells and gene set enrichment analysis for hub genes suggested that there were some associations between these genes and the pathogenesis of CRC. CONCLUSION: This study identified modules that were significantly associated with CRC, four novel hub genes, and further analysis of these genes. This may provide a little new insights and directions into the potential pathogenesis of CRC.

Identification of Robust and Key Differentially Expressed Genes during C2C12 Cell Myogenesis Based on Multiomics Data.

Myogenesis is a central step in prenatal myofiber formation, postnatal myofiber hypertrophy, and muscle damage repair in adulthood. RNA-Seq technology has greatly helped reveal the molecular mechanism of myogenesis, but batch effects in different experiments inevitably lead to misinterpretation of differentially expressed genes (DEGs). We previously applied the robust rank aggregation (RRA) method to effectively circumvent batch effects across multiple RNA-Seq datasets from 3T3-L1 cells. Here, we also used the RRA method to integrate nine RNA-Seq datasets from C2C12 cells and obtained 3140 robust DEGs between myoblasts and myotubes, which were then validated with array expression profiles and H3K27ac signals. The upregulated robust DEGs were highly enriched in gene ontology (GO) terms related to muscle cell differentiation and development. Considering that the cooperative binding of transcription factors (TFs) to enhancers to regulate downstream gene expression is a classical epigenetic mechanism, differentially expressed TFs (DETFs) were screened, and potential novel myogenic factors (MAF, BCL6, and ESR1) with high connection degree in protein-protein interaction (PPI) network were presented. Moreover, KLF5 cooperatively binds with the three key myogenic factors (MYOD, MYOG, and MEF2D) in C2C12 cells. Motif analysis speculates that the binding of MYOD and MYOG is KLF5-independent, while MEF2D is KLF5-dependent. It was revealed that KLF5-binding sites could be exploited to filter redundant MYOD-, MYOG-, and MEF2D-binding sites to focus on key enhancers for myogenesis. Further functional annotation of KLF5-binding sites suggested that KLF5 may regulate myogenesis through the PI3K-AKt signaling pathway, Rap1 signaling pathway, and the Hippo signaling pathway. In general, our study provides a wealth of untapped candidate targets for myogenesis and contributes new insights into the core regulatory mechanisms of myogenesis relying on KLF5-binding signal.

Prediction of potential molecular markers of bovine mastitis by meta-analysis of differentially expressed genes using combined p value and robust rank aggregation.

Bovine mastitis causes significant economic loss to the dairy industry by affecting milk quality and quantity. Escherichia coli and Staphylococcus aureus are the two common mastitis-causing bacteria among the consortia of mastitis pathogens, wherein E. coli is an opportunistic environmental pathogen, and S. aureus is a contagious pathogen. This study was designed to predict molecular markers of bovine mastitis by meta-analysis of differentially expressed genes (DEG) in E. coli- or S. aureus-infected mammary epithelial cells (MECs) using p value combination and robust rank aggregation (RRA) methods. High-throughput transcriptome of bovine MECs, infected with E. coli or S. aureus, were analyzed, and correlation of z-scores were computed for the expression datasets to identify the lineage profile and functional ontology of DEGs. Key pathways enriched in infected MECs were deciphered by Gene Set Enrichment Analysis (GSEA), following which combined p value and RRA were used to perform DEG meta-analysis to limit type I error in the analysis. The miRNA-gene networks were then built to uncover potential molecular markers of mastitis. Lineage profiling of MECs showed that the gene expression levels were associated with mammary tissue lineage. The up-regulated genes were enriched in immune-related pathways, whereas down-regulated genes influenced the cellular processes. GSEA analysis of DEGs deciphered the involvement of Toll-like receptor (TLR), and NF-kappa B signaling pathway during infection. Comparison after meta-analysis yielded with genes ZC3H12A, RND1, and MAP3K8 having significant expression levels in both E. coli and S. aureus dataset, and on evaluating miRNA-gene network, 7 pairs were common to both sets identifying them as potential molecular markers.

Integrated analysis of multiple microarray studies to identify potential pathogenic gene modules in preeclampsia.

BACKGROUND: Preeclampsia is a life-threatening hypertensive disorder during pregnancy, while underlying pathogenesis and its diagnosis are incomplete. METHODS: In this study, we utilized the Robust Rank Aggregation method to integrate 6 eligible preeclampsia microarray datasets from Gene Expression Omnibus database. We used linear regression to assess the associations between significant differentially expressed genes (DEGs) and blood pressure. Functional annotation, protein-protein interaction, Gene Set Enrichment Analysis (GSEA) and single sample GSEA were employed for investigating underlying pathogenesis in preeclampsia. RESULTS: We filtered 52 DEGs and further screened for 5 hub genes (leptin, pappalysin 2, endoglin, fms related receptor tyrosine kinase 1, tripartite motif containing 24) that were positively correlated with both systolic blood pressure and diastolic blood pressure. Receiver operating characteristic indicated that hub genes were potential biomarkers for diagnosis and prognosis in preeclampsia. GSEA for single hub gene revealed that they were all closely related to angiogenesis and estrogen response in preeclampsia. Moreover, single sample GSEA showed that the expression levels of 5 hub genes were correlated with those of immune cells in immunologic microenvironment at maternal-fetal interface. CONCLUSIONS: These findings provide new insights into underlying pathogenesis in preeclampsia; 5 hub genes were identified as biomarkers for diagnosis and prognosis in preeclampsia.

An Adaptive Rank Aggregation-Based Ensemble Multi-Filter Feature Selection Method in Software Defect Prediction.

Feature selection is known to be an applicable solution to address the problem of high dimensionality in software defect prediction (SDP). However, choosing an appropriate filter feature selection (FFS) method that will generate and guarantee optimal features in SDP is an open research issue, known as the filter rank selection problem. As a solution, the combination of multiple filter methods can alleviate the filter rank selection problem. In this study, a novel adaptive rank aggregation-based ensemble multi-filter feature selection (AREMFFS) method is proposed to resolve high dimensionality and filter rank selection problems in SDP. Specifically, the proposed AREMFFS method is based on assessing and combining the strengths of individual FFS methods by aggregating multiple rank lists in the generation and subsequent selection of top-ranked features to be used in the SDP process. The efficacy of the proposed AREMFFS method is evaluated with decision tree (DT) and naïve Bayes (NB) models on defect datasets from different repositories with diverse defect granularities. Findings from the experimental results indicated the superiority of AREMFFS over other baseline FFS methods that were evaluated, existing rank aggregation based multi-filter FS methods, and variants of AREMFFS as developed in this study. That is, the proposed AREMFFS method not only had a superior effect on prediction performances of SDP models but also outperformed baseline FS methods and existing rank aggregation based multi-filter FS methods. Therefore, this study recommends the combination of multiple FFS methods to utilize the strength of respective FFS methods and take advantage of filter-filter relationships in selecting optimal features for SDP processes.

Identification and validation of reference genes for qRT-PCR based studies in horse gram (Macrotyloma uniflorum).

The quantitative real-time polymerase chain reaction (qRT-PCR) is the most sensitive and commonly used technique for gene expression studies in biological systems. However, the reliability of qRT-PCR results depends on the selection of reference gene(s) for data normalization. Horse gram (Macrotyloma uniflorum) is an important legume crop on which several molecular studies have been reported. However, the stability of reference genes has not been evaluated. In the present study, nine candidate reference genes were identified from horse gram RNA-seq data and evaluated in two horse gram genotypes, HPK4 and HPKM317 under six abiotic stresses viz. cold, drought, salinity, heat, abscisic acid and methyl viologen-induced oxidative stress. The results were evaluated using geNorm, Bestkeeper, Normfinder and delta-delta Ct methods and comprehensive ranking was assigned using RefFinder and RankAggreg software. The overall result showed that TCTP was one of the most stable genes in all samples and in genotype HPK4, while in HPKM317 profilin was most stably expressed. However, PSMA5 was identified as least stable in all the experimental conditions. Expression of target genes dehydrin and early response to dehydration 6 under drought stress was also validated using TCTP and profilin for data normalization, either alone or in combination, which confirmed their suitability for qRT-PCR data normalization. Thus, TCTP and profilin genes may be used for qRT-PCR data normalization for molecular and genomic studies in horse gram. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s12298-021-01104-0.

Identification of hub genes in papillary thyroid carcinoma: robust rank aggregation and weighted gene co-expression network analysis.

BACKGROUND: Papillary thyroid carcinoma (PTC), which is the most common endocrine malignancy, has been steadily increasing worldwide in incidence over the years, while mechanisms underlying the pathogenesis and diagnostic for PTC are incomplete. The purpose of this study is to identify potential biomarkers for diagnosis of PTC, and provide new insights into pathogenesis of PTC. METHODS: Based on weighted gene co-expression network analysis, Robust Rank Aggregation, functional annotation, GSEA and DNA methylation, were employed for investigating potential biomarkers for diagnosis of PTC. RESULTS: Black and turquoise modules were identified in the gene co-expression network constructed by 1807 DEGs that from 6 eligible gene expression profiles of Gene Expression Omnibus database based on Robust Rank Aggregation and weighted gene co-expression network analysis. Hub genes were significantly down-regulated and the expression levels of the hub genes were different in different stages in hub gene verification. ROC curves indicated all hub genes had good diagnostic value for PTC (except for ABCA6 AUC = 89.5%, the 15 genes with AUC > 90%). Methylation analysis showed that hub gene verification ABCA6, ACACB, RMDN1 and TFPI were identified as differentially methylated genes, and the decreased expression level of these genes may relate to abnormal DNA methylation. Moreover, the expression levels of 8 top hub genes were correlated with tumor purity and tumor-infiltrating immune cells. These findings, including functional annotations and GSEA provide new insights into pathogenesis of PTC. CONCLUSIONS: The hub genes and methylation of hub genes may as potential biomarkers provide new insights for diagnosis of PTC, and all these findings may be the direction to study the mechanisms underlying of PTC in the future.

Dysregulated microRNAs in laryngeal cancer: a comprehensive meta-analysis using a robust rank aggregation approach.

We need a reasonable method of compiling data from different studies regarding the expression of microRNA (miRNA) in laryngeal squamous cell carcinoma (LSCC). The robust rank aggregation method was used to integrate the rank lists of miRNAs from 11 studies. The enrichment analysis was performed on target genes of meta-signature miRNAs. The Cancer Genome Atlas database was used to confirm the results of meta-analysis. Three meta-signature miRNAs (miR-21-5p, miR-196a-5p and miR-145-5p) were obtained. All three miRNAs could be prognostic for LSCC. The enrichment analysis showed that these miRNAs were associated significantly with multiple cancer-related signaling pathways. The robust rank aggregation approach is an effective way to identify important miRNAs from different studies. All identified miRNAs could be candidates for LSCC diagnostic and prognostic biomarkers.

Selection of the optimal personalized treatment from multiple treatments with multivariate outcome measures.

In this work, we propose a novel method for individualized treatment selection when the treatment response is multivariate. For the K treatment (K ≥2) scenario we compare quantities that are suitable indexes based on outcome variables for each treatment conditional on patient-specific scores constructed from collected covariate measurements. Our method covers any number of treatments and outcome variables, and it can be applied for a broad set of models. The proposed method uses a rank aggregation technique to estimate an ordering of treatments based on ranked lists of treatment performance measures such as smooth conditional means and conditional probability of a response for one treatment dominating others. The method has the flexibility to incorporate patient and clinician preferences to the optimal treatment decision on an individual case basis. A simulation study demonstrates the performance of the proposed method in finite samples. We also present data analyses using HIV and Diabetes clinical trials data to show the applicability of the proposed procedure for real data.

GPS: Identification of disease genes by rank aggregation of multi-genomic scoring schemes.

In solving the gene prioritization problem, ranking candidate genes from most to least promising is attempted before further experimental validation. Integrating the results of various data sources and methods tends to result in a better performance when solving the gene prioritization problem. Therefore, a wide range of datasets and algorithms was investigated; these included topological features of protein networks, physicochemical characteristics and blast similarity scores of protein sequences, gene ontology, biological pathways, and tissue-based data sources. The novelty of this study lies in how the best-performing methods and reliable multi-genomic data sources were applied in an efficient two-step approach. In the first step, various multi-genomic data sources and algorithms were evaluated and seven best-performing rankers were then applied to prioritize candidate genes in different ways. In the second step, global prioritization was obtained by aggregating several scoring schemes. The results showed that protein networks, functional linkage networks, gene ontology, and biological pathway data sources have a significant impact on the quality of the gene prioritization approach. The findings also demonstrated a direct relationship between the degree of genes and the ranking quality of the evaluated tools. This approach outperformed previously published algorithms (e.g., DIR, GPEC, GeneDistiller, and Endeavour) in all evaluation metrices and led to the development of GPS software. Its user-friendly interface and accuracy makes GPS a powerful tool for the identification of human disease genes. GPS is available at http://gpsranker.com and http://LBB.ut.ac.ir.

Identification of key genes and pathways of thyroid cancer by integrated bioinformatics analysis.

Thyroid cancer is a common endocrine malignancy with a rapidly increasing incidence worldwide. Although its mortality is steady or declining because of earlier diagnoses, its survival rate varies because of different tumour types. Thus, the aim of this study was to identify key biomarkers and novel therapeutic targets in thyroid cancer. The expression profiles of GSE3467, GSE5364, GSE29265 and GSE53157 were downloaded from the Gene Expression Omnibus database, which included a total of 97 thyroid cancer and 48 normal samples. After screening significant differentially expressed genes (DEGs) in each data set, we used the robust rank aggregation method to identify 358 robust DEGs, including 135 upregulated and 224 downregulated genes, in four datasets. Gene Ontology and Kyoto Encyclopaedia of Genes and Genomes pathway enrichment analyses of DEGs were performed by DAVID and the KOBAS online database, respectively. The results showed that these DEGs were significantly enriched in various cancer-related functions and pathways. Then, the STRING database was used to construct the protein-protein interaction network, and modules analysis was performed. Finally, we filtered out five hub genes, including LPAR5, NMU, FN1, NPY1R, and CXCL12, from the whole network. Expression validation and survival analysis of these hub genes based on the The Cancer Genome Atlas database suggested the robustness of the above results. In conclusion, these results provided novel and reliable biomarkers for thyroid cancer, which will be useful for further clinical applications in thyroid cancer diagnosis, prognosis and targeted therapy.