Human gene correlation analysis (HGCA): a tool for the identification of transcriptionally co-expressed genes

Activation of the FOXM1/ASF1B/PRDX3 axis confers hyperproliferative and antioxidative stress reactivity to gastric cancer

Nucleosome assembly during DNA replication is dependent on histone chaperones. Recent studies suggest that dysregulated histone chaperones contribute to cancer progression, including gastric cancer (GC). Further studies are required to explore the prognostic and therapeutic implications of histone chaperones and their mechanisms of action in GC progression. Here we identified histone chaperone ASF1B as a potential biomarker for GC proliferation and prognosis. ASF1B was significantly upregulated in GC, which was associated with poor prognosis. In vitro and in vivo experiments demonstrated that the inhibition of ASF1B suppressed the malignant characteristics of GC, while overexpression of ASF1B had the opposite effect. Mechanistically, transcription factor FOXM1 directly bound to the ASF1B-promoter region, thereby regulating its transcription. Treatment with thiostrepton, a FOXM1 inhibitor, not only suppressed ASF1B expression, but also inhibited GC progression. Furthermore, ASF1B regulated the mitochondrial protein peroxiredoxin 3 (PRDX3) transcription in a FOXM1-dependent manner. The crucial role of ASF1B-regulated PRDX3 in GC cell proliferation and oxidative stress balance was also elucidated. In summary, our study suggests that the FOXM1-ASF1B-PRDX3 axis is a potential therapeutic target for treating GC.

HGCA2.0: An RNA-Seq Based Webtool for Gene Coexpression Analysis in Homo sapiens

Genes with similar expression patterns in a set of diverse samples may be considered coexpressed. Human Gene Coexpression Analysis 2.0 (HGCA2.0) is a webtool which studies the global coexpression landscape of human genes. The website is based on the hierarchical clustering of 55,431 Homo sapiens genes based on a large-scale coexpression analysis of 3500 GTEx bulk RNA-Seq samples of healthy individuals, which were selected as the best representative samples of each tissue type. HGCA2.0 presents subclades of coexpressed genes to a gene of interest, and performs various built-in gene term enrichment analyses on the coexpressed genes, including gene ontologies, biological pathways, protein families, and diseases, while also being unique in revealing enriched transcription factors driving coexpression. HGCA2.0 has been successful in identifying not only genes with ubiquitous expression patterns, but also tissue-specific genes. Benchmarking showed that HGCA2.0 belongs to the top performing coexpression webtools, as shown by STRING analysis. HGCA2.0 creates working hypotheses for the discovery of gene partners or common biological processes that can be experimentally validated. It offers a simple and intuitive website design and user interface, as well as an API endpoint.

Density-based detection of cell transition states to construct disparate and bifurcating trajectories

Tree- and linear-shaped cell differentiation trajectories have been widely observed in developmental biologies and can be also inferred through computational methods from single-cell RNA-sequencing datasets. However, trajectories with complicated topologies such as loops, disparate lineages and bifurcating hierarchy remain difficult to infer accurately. Here, we introduce a density-based trajectory inference method capable of constructing diverse shapes of topological patterns including the most intriguing bifurcations. The novelty of our method is a step to exploit overlapping probability distributions to identify transition states of cells for determining connectability between cell clusters, and another step to infer a stable trajectory through a base-topology guided iterative fitting. Our method precisely re-constructed various benchmark reference trajectories. As a case study to demonstrate practical usefulness, our method was tested on single-cell RNA sequencing profiles of blood cells of SARS-CoV-2-infected patients. We not only re-discovered the linear trajectory bridging the transition from IgM plasmablast cells to developing neutrophils, and also found a previously-undiscovered lineage which can be rigorously supported by differentially expressed gene analysis.

Analysis of ROQUIN, Tristetraprolin (TTP), and BDNF/miR-16/TTP regulatory axis in late onset Alzheimer’s disease

Alzheimer’s disease (AD) is a heterogeneous degenerative disorder of the brain that is on the rise worldwide. One of the critical processes that might be disturbed in AD is gene expression regulation. Tristetraprolin (TTP) and RC3H1 gene (ROQUIN) are two RNA-binding proteins (RBPs) that target AU-rich elements (AREs) and constitutive decay elements (CDEs), respectively. TTP and ROQUIN, members of the CCCH zinc-finger protein family, have been demonstrated to fine-tune numerous inflammatory factors. In addition, miR-16 has distinct characteristics and may influence the target mRNA through the ARE site. Interestingly, BDNF mRNA has ARE sites in the 3’ untranslated region (UTR) and can be targeted by regulatory factors, such as TTP and miR-16 on MRE sequences, forming BDNF/miR-16/TTP regulatory axis. A number of two microarray datasets were downloaded, including information on mRNAs (GSE106241) and miRNAs (GSE157239) from individuals with AD and corresponding controls. R software was used to identify BDNF, TTP, ROQUIN, and miR-16 expression levels in temporal cortex (TC) tissue datasets. Q-PCR was also used to evaluate the expression of these regulatory factors and the expression of BDNF in the blood of 50 patients with AD and 50 controls. Bioinformatic evaluation showed that TTP and miR-16 overexpression might act as post-transcriptional regulatory factors to control BDNF expression in AD in TC samples. Instead, this expression pattern was not found in peripheral blood samples from patients with AD compared to normal controls. ROQUIN expression was increased in the peripheral blood of patients with AD. Hsa-miR-16-5p levels did not show significant differences in peripheral blood samples. Finally, it was shown that TTP and BDNF, based on evaluating the receiver operating characteristic (ROC), effectively identify patients with AD from healthy controls. This study could provide a new perspective on the molecular regulatory processes associated with AD pathogenic mechanisms linked to the BDNF growth factor, although further research is needed on the possible roles of these factors in AD.

Approaches in Gene Coexpression Analysis in Eukaryotes

Simple Summary

Genes whose expression levels rise and fall similarly in a large set of samples, may be considered coexpressed. Gene coexpression analysis refers to the en masse discovery of coexpressed genes from a large variety of transcriptomic experiments. The type of biological networks that studies gene coexpression, known as Gene Coexpression Networks, consist of an undirected graph depicting genes and their coexpression relationships. Coexpressed genes are clustered in smaller subnetworks, the predominant biological roles of which can be determined through enrichment analysis. By studying well-annotated gene partners, the attribution of new roles to genes of unknown function or assumption for participation in common metabolic pathways can be achieved, through a guilt-by-association approach. In this review, we present key issues in gene coexpression analysis, as well as the most popular tools that perform it.

Abstract

Gene coexpression analysis constitutes a widely used practice for gene partner identification and gene function prediction, consisting of many intricate procedures. The analysis begins with the collection of primary transcriptomic data and their preprocessing, continues with the calculation of the similarity between genes based on their expression values in the selected sample dataset and results in the construction and visualisation of a gene coexpression network (GCN) and its evaluation using biological term enrichment analysis. As gene coexpression analysis has been studied extensively, we present most parts of the methodology in a clear manner and the reasoning behind the selection of some of the techniques. In this review, we offer a comprehensive and comprehensible account of the steps required for performing a complete gene coexpression analysis in eukaryotic organisms. We comment on the use of RNA-Seq vs. microarrays, as well as the best practices for GCN construction. Furthermore, we recount the most popular webtools and standalone applications performing gene coexpression analysis, with details on their methods, features and outputs.

Arabidopsis Coexpression Tool: a tool for gene coexpression analysis in Arabidopsis thaliana

Gene coexpression analysis refers to the discovery of sets of genes which exhibit similar expression patterns across multiple transcriptomic data sets, such as microarray experiment data of public repositories. Arabidopsis Coexpression Tool (ACT), a gene coexpression analysis web tool for Arabidopsis thaliana, identifies genes which are correlated to a driver gene. Primary microarray data from ATH1 Affymetrix platform were processed with Single-Channel Array Normalization algorithm and combined to produce a coexpression tree which contains ∼21,000 A. thaliana genes. ACT was developed to present subclades of coexpressed genes, as well as to perform gene set enrichment analysis, being unique in revealing enriched transcription factors targeting coexpressed genes. ACT offers a simple and user-friendly interface producing working hypotheses which can be experimentally verified for the discovery of gene partnership, pathway membership, and transcriptional regulation. ACT analyses have been successful in identifying not only genes with coordinated ubiquitous expressions but also genes with tissue-specific expressions.

Biomolecule and Bioentity Interaction Databases in Systems Biology: A Comprehensive Review

Technological advances in high-throughput techniques have resulted in tremendous growth of complex biological datasets providing evidence regarding various biomolecular interactions. To cope with this data flood, computational approaches, web services, and databases have been implemented to deal with issues such as data integration, visualization, exploration, organization, scalability, and complexity. Nevertheless, as the number of such sets increases, it is becoming more and more difficult for an end user to know what the scope and focus of each repository is and how redundant the information between them is. Several repositories have a more general scope, while others focus on specialized aspects, such as specific organisms or biological systems. Unfortunately, many of these databases are self-contained or poorly documented and maintained. For a clearer view, in this article we provide a comprehensive categorization, comparison and evaluation of such repositories for different bioentity interaction types. We discuss most of the publicly available services based on their content, sources of information, data representation methods, user-friendliness, scope and interconnectivity, and we comment on their strengths and weaknesses. We aim for this review to reach a broad readership varying from biomedical beginners to experts and serve as a reference article in the field of Network Biology.

A variable selection approach for highly correlated predictors in high-dimensional genomic data

MOTIVATION

In genomic studies, identifying biomarkers associated with a variable of interest is a major concern in biomedical research. Regularized approaches are classically used to perform variable selection in high-dimensional linear models. However, these methods can fail in highly correlated settings.

RESULTS

We propose a novel variable selection approach called WLasso, taking these correlations into account. It consists in rewriting the initial high-dimensional linear model to remove the correlation between the biomarkers (predictors) and in applying the generalized Lasso criterion. The performance of WLasso is assessed using synthetic data in several scenarios and compared with recent alternative approaches. The results show that when the biomarkers are highly correlated, WLasso outperforms the other approaches in sparse high-dimensional frameworks. The method is also illustrated on publicly available gene expression data in breast cancer.

AVAILABILITY

Our method is implemented in the WLasso R package which is available from the Comprehensive R Archive Network (CRAN).

SUPPLEMENTARY INFORMATION

Supplementary material is available at Bioinformatics online.

Transcriptome of Edwardsiella anguillarum in vivo and in vitro revealed two-component system, ABC transporter and flagellar assembly are three pathways pathogenic to European eel (Anguilla anguilla)

Edwardsiella anguillarum is one of the common bacterial pathogens for the cultivated eels in China. The aim of this study was to reveal the cause of E. anguillarum pathogenic to European eel (Anguilla anguilla) from the perspective of the transcriptome. In this study, we first prepared E. anguillarum cultured in vitro and analysed the whole transcriptome after extracting the total RNA. Then, eels were i.p injected with E. anguillarum, and total RNA were extracted from the liver of European eels 48 h after the infection. After sequencing the transcriptome, we obtained average 1.97 × 10⁸ clean reads cultured in vitro and 1.36 × 10⁵ clean reads located in vivo after annotating all reads into the genome of E. anguillarum. The whole transcriptome showed, compared to the E. anguillarum cultured in vitro, 503 significantly up and 657 significantly down-regulated different expressed genes (DEGs) were observed. KEGG analysis showed that 38 DEGs of Two-Component System, 41 DEGs of ABC transporter, and 10 DEGs flagellar assembly pathways were highly upregulated in E. anguillarum located in vivo. Then, we designed primers to analyse the up-regulated DEGs through qRT-PCR and confirmed some up-regulated DEGs. The results of this study provide important reference for the further study of pathogen-host interaction between E. anguillarum and European eel.

A Deep Learning Approach Validates Genetic Risk Factors for Late Toxicity After Prostate Cancer Radiotherapy in a REQUITE Multi-National Cohort

Background: REQUITE (validating pREdictive models and biomarkers of radiotherapy toxicity to reduce side effects and improve QUalITy of lifE in cancer survivors) is an international prospective cohort study. The purpose of this project was to analyse a cohort of patients recruited into REQUITE using a deep learning algorithm to identify patient-specific features associated with the development of toxicity, and test the approach by attempting to validate previously published genetic risk factors. Methods: The study involved REQUITE prostate cancer patients treated with external beam radiotherapy who had complete 2-year follow-up. We used five separate late toxicity endpoints: ≥grade 1 late rectal bleeding, ≥grade 2 urinary frequency, ≥grade 1 haematuria, ≥ grade 2 nocturia, ≥ grade 1 decreased urinary stream. Forty-three single nucleotide polymorphisms (SNPs) already reported in the literature to be associated with the toxicity endpoints were included in the analysis. No SNP had been studied before in the REQUITE cohort. Deep Sparse AutoEncoders (DSAE) were trained to recognize features (SNPs) identifying patients with no toxicity and tested on a different independent mixed population including patients without and with toxicity. Results: One thousand, four hundred and one patients were included, and toxicity rates were: rectal bleeding 11.7%, urinary frequency 4%, haematuria 5.5%, nocturia 7.8%, decreased urinary stream 17.1%. Twenty-four of the 43 SNPs that were associated with the toxicity endpoints were validated as identifying patients with toxicity. Twenty of the 24 SNPs were associated with the same toxicity endpoint as reported in the literature: 9 SNPs for urinary symptoms and 11 SNPs for overall toxicity. The other 4 SNPs were associated with a different endpoint. Conclusion: Deep learning algorithms can validate SNPs associated with toxicity after radiotherapy for prostate cancer. The method should be studied further to identify polygenic SNP risk signatures for radiotherapy toxicity. The signatures could then be included in integrated normal tissue complication probability models and tested for their ability to personalize radiotherapy treatment planning.

MyoMiner: explore gene co-expression in normal and pathological muscle

BACKGROUND

High-throughput transcriptomics measures mRNA levels for thousands of genes in a biological sample. Most gene expression studies aim to identify genes that are differentially expressed between different biological conditions, such as between healthy and diseased states. However, these data can also be used to identify genes that are co-expressed within a biological condition. Gene co-expression is used in a guilt-by-association approach to prioritize candidate genes that could be involved in disease, and to gain insights into the functions of genes, protein relations, and signaling pathways. Most existing gene co-expression databases are generic, amalgamating data for a given organism regardless of tissue-type.

METHODS

To study muscle-specific gene co-expression in both normal and pathological states, publicly available gene expression data were acquired for 2376 mouse and 2228 human striated muscle samples, and separated into 142 categories based on species (human or mouse), tissue origin, age, gender, anatomic part, and experimental condition. Co-expression values were calculated for each category to create the MyoMiner database.

RESULTS

Within each category, users can select a gene of interest, and the MyoMiner web interface will return all correlated genes. For each co-expressed gene pair, adjusted p-value and confidence intervals are provided as measures of expression correlation strength. A standardized expression-level scatterplot is available for every gene pair r-value. MyoMiner has two extra functions: (a) a network interface for creating a 2-shell correlation network, based either on the most highly correlated genes or from a list of genes provided by the user with the option to include linked genes from the database and (b) a comparison tool from which the users can test whether any two correlation coefficients from different conditions are significantly different.

CONCLUSIONS

These co-expression analyses will help investigators to delineate the tissue-, cell-, and pathology-specific elements of muscle protein interactions, cell signaling and gene regulation. Changes in co-expression between pathologic and healthy tissue may suggest new disease mechanisms and help define novel therapeutic targets. Thus, MyoMiner is a powerful muscle-specific database for the discovery of genes that are associated with related functions based on their co-expression. MyoMiner is freely available at https://www.sys-myo.com/myominer.

Precision Lasso: accounting for correlations and linear dependencies in high-dimensional genomic data

MOTIVATION

Association studies to discover links between genetic markers and phenotypes are central to bioinformatics. Methods of regularized regression, such as variants of the Lasso, are popular for this task. Despite the good predictive performance of these methods in the average case, they suffer from unstable selections of correlated variables and inconsistent selections of linearly dependent variables. Unfortunately, as we demonstrate empirically, such problematic situations of correlated and linearly dependent variables often exist in genomic datasets and lead to under-performance of classical methods of variable selection.

RESULTS

To address these challenges, we propose the Precision Lasso. Precision Lasso is a Lasso variant that promotes sparse variable selection by regularization governed by the covariance and inverse covariance matrices of explanatory variables. We illustrate its capacity for stable and consistent variable selection in simulated data with highly correlated and linearly dependent variables. We then demonstrate the effectiveness of the Precision Lasso to select meaningful variables from transcriptomic profiles of breast cancer patients. Our results indicate that in settings with correlated and linearly dependent variables, the Precision Lasso outperforms popular methods of variable selection such as the Lasso, the Elastic Net and Minimax Concave Penalty (MCP) regression.

AVAILABILITY AND IMPLEMENTATION

Software is available at https://github.com/HaohanWang/thePrecisionLasso.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Integration of Machine Learning Methods to Dissect Genetically Imputed Transcriptomic Profiles in Alzheimer's Disease

The genetic component of many common traits is associated with the gene expression and several variants act as expression quantitative loci, regulating the gene expression in a tissue specific manner. In this work, we applied tissue-specific cis-eQTL gene expression prediction models on the genotype of 808 samples including controls, subjects with mild cognitive impairment, and patients with Alzheimer's Disease. We then dissected the imputed transcriptomic profiles by means of different unsupervised and supervised machine learning approaches to identify potential biological associations. Our analysis suggests that unsupervised and supervised methods can provide complementary information, which can be integrated for a better characterization of the underlying biological system. In particular, a variational autoencoder representation of the transcriptomic profiles, followed by a support vector machine classification, has been used for tissue-specific gene prioritizations. Interestingly, the achieved gene prioritizations can be efficiently integrated as a feature selection step for improving the accuracy of deep learning classifier networks. The identified gene-tissue information suggests a potential role for inflammatory and regulatory processes in gut-brain axis related tissues. In line with the expected low heritability that can be apportioned to eQTL variants, we were able to achieve only relatively low prediction capability with deep learning classification models. However, our analysis revealed that the classification power strongly depends on the network structure, with recurrent neural networks being the best performing network class. Interestingly, cross-tissue analysis suggests a potentially greater role of models trained in brain tissues also by considering dementia-related endophenotypes. Overall, the present analysis suggests that the combination of supervised and unsupervised machine learning techniques can be used for the evaluation of high dimensional omics data.

Reconstruction and inference of the Lactococcus lactis MG1363 gene co-expression network

Lactic acid bacteria are Gram-positive bacteria used throughout the world in many industrial applications for their acidification, flavor and texture formation attributes. One of the species, Lactococcus lactis, is employed for the production of fermented milk products like cheese, buttermilk and quark. It ferments lactose to lactic acid and, thus, helps improve the shelf life of the products. Many physiological and transcriptome studies have been performed in L. lactis in order to comprehend and improve its biotechnological assets. Using large amounts of transcriptome data to understand and predict the behavior of biological processes in bacterial or other cell types is a complex task. Gene networks enable predicting gene behavior and function in the context of transcriptionally linked processes. We reconstruct and present the gene co-expression network (GCN) for the most widely studied L. lactis strain, MG1363, using publicly available transcriptome data. Several methods exist to generate and judge the quality of GCNs. Different reconstruction methods lead to networks with varying structural properties, consequently altering gene clusters. We compared the structural properties of the MG1363 GCNs generated by five methods, namely Pearson correlation, Spearman correlation, GeneNet, Weighted Gene Co-expression Network Analysis (WGCNA), and Sparse PArtial Correlation Estimation (SPACE). Using SPACE, we generated an L. lactis MG1363 GCN and assessed its quality using modularity and structural and biological criteria. The L. lactis MG1363 GCN has structural properties similar to those of the gold-standard networks of Escherichia coli K-12 and Bacillus subtilis 168. We showcase that the network can be used to mine for genes with similar expression profiles that are also generally linked to the same biological process.

Dynamic sporulation gene co-expression networks for Bacillus subtilis 168 and the food-borne isolate Bacillus amyloliquefaciens: a transcriptomic model

Sporulation is a survival strategy, adapted by bacterial cells in response to harsh environmental adversities. The adaptation potential differs between strains and the variations may arise from differences in gene regulation. Gene networks are a valuable way of studying such regulation processes and establishing associations between genes. We reconstructed and compared sporulation gene co-expression networks (GCNs) of the model laboratory strain Bacillus subtilis 168 and the food-borne industrial isolate Bacillus amyloliquefaciens. Transcriptome data obtained from samples of six stages during the sporulation process were used for network inference. Subsequently, a gene set enrichment analysis was performed to compare the reconstructed GCNs of B. subtilis 168 and B. amyloliquefaciens with respect to biological functions, which showed the enriched modules with coherent functional groups associated with sporulation. On basis of the GCNs and time-evolution of differentially expressed genes, we could identify novel candidate genes strongly associated with sporulation in B. subtilis 168 and B. amyloliquefaciens. The GCNs offer a framework for exploring transcription factors, their targets, and co-expressed genes during sporulation. Furthermore, the methodology described here can conveniently be applied to other species or biological processes.

Virtual Sorting Has a Distinctive Advantage in Identification of Anticorrelated Genes and Further Negative Regulators of Immune Cell Subpopulations

Immune cells are highly plastic in both gene expression and cell phenotype. We have established a method of gene expressional plasticity and virtual sorting to evaluate immune cell subpopulations and their characteristic genes in human CD4⁺ T cells. In this study, we continued to investigate the informatics mechanism on the effectiveness of virtual sorting. We found that virtual sorting had an overall positive correlation to the Pearson correlation in the identification of positively correlated genes. However, owing to nonlinear biological anticorrelation, virtual sorting showed a distinctive advantage for anticorrelated genes, suggesting an important role in the identification of negative regulators. In addition, based on virtual sorting results, we identified two basic gene sets among highly plastic genes, i.e., highly plastic cell cycle-associated molecules and highly plastic immune and defense response-associated molecules. Genes within each set tended to be positively connected, but genes between two sets were often anticorrelated. Further analysis revealed preferential transcription factor binding motifs existed between highly plastic cell cycle-associated molecules and highly plastic immune and defense response-associated molecules. Our results strongly suggested predetermined regulation, which was called an immune cell internal phenotype, should exist and could be mined by virtual sorting analysis. This provided efficient functional clues to study immune cell phenotypes and their regulation. Moreover, the current substantial virtual sorting results in both CD4⁺ T cells and B cells provide a useful resource for big-data-driven experimental studies and knowledge discoveries.

Parkinson's disease-associated genetic variation is linked to quantitative expression of inflammatory genes

Genome-wide association studies (GWAS) have linked dozens of single nucleotide polymorphisms (SNPs) with Parkinson’s disease (PD) risk. Ascertaining the functional and eventual causal mechanisms underlying these relationships has proven difficult. The majority of risk SNPs, and nearby SNPs in linkage disequilibrium (LD), are found in intergenic or intronic regions and confer risk through allele-dependent expression of multiple unknown target genes. Combining GWAS results with publicly available GTEx data, generated through eQTL (expression quantitative trait loci) identification studies, enables a direct association of SNPs to gene expression levels and aids in narrowing the large population of potential genetic targets for hypothesis-driven experimental cell biology. Separately, overlapping of SNPs with putative enhancer segmentations can strengthen target filtering. We report here the results of analyzing 7,607 PD risk SNPs along with an additional 23,759 high linkage disequilibrium-associated variants paired with eQTL gene expression. We found that enrichment analysis on the set of genes following target filtering pointed to a single large LD block at 6p21 that contained multiple HLA-MHC-II genes. These MHC-II genes remain associated with PD when the genes were filtered for correlation between GWAS significance and eQTL levels, strongly indicating a direct effect on PD etiology.

Cogena, a novel tool for co-expressed gene-set enrichment analysis, applied to drug repositioning and drug mode of action discovery

BACKGROUND

Drug repositioning, finding new indications for existing drugs, has gained much recent attention as a potentially efficient and economical strategy for accelerating new therapies into the clinic. Although improvement in the sensitivity of computational drug repositioning methods has identified numerous credible repositioning opportunities, few have been progressed. Arguably the "black box" nature of drug action in a new indication is one of the main blocks to progression, highlighting the need for methods that inform on the broader target mechanism in the disease context.

RESULTS

We demonstrate that the analysis of co-expressed genes may be a critical first step towards illumination of both disease pathology and mode of drug action. We achieve this using a novel framework, co-expressed gene-set enrichment analysis (cogena) for co-expression analysis of gene expression signatures and gene set enrichment analysis of co-expressed genes. The cogena framework enables simultaneous, pathway driven, disease and drug repositioning analysis. Cogena can be used to illuminate coordinated changes within disease transcriptomes and identify drugs acting mechanistically within this framework. We illustrate this using a psoriatic skin transcriptome, as an exemplar, and recover two widely used Psoriasis drugs (Methotrexate and Ciclosporin) with distinct modes of action. Cogena out-performs the results of Connectivity Map and NFFinder webservers in similar disease transcriptome analyses. Furthermore, we investigated the literature support for the other top-ranked compounds to treat psoriasis and showed how the outputs of cogena analysis can contribute new insight to support the progression of drugs into the clinic. We have made cogena freely available within Bioconductor or https://github.com/zhilongjia/cogena .

CONCLUSIONS

In conclusion, by targeting co-expressed genes within disease transcriptomes, cogena offers novel biological insight, which can be effectively harnessed for drug discovery and repositioning, allowing the grouping and prioritisation of drug repositioning candidates on the basis of putative mode of action.

Immuno-Navigator, a batch-corrected coexpression database, reveals cell type-specific gene networks in the immune system

High-throughput gene expression data are one of the primary resources for exploring complex intracellular dynamics in modern biology. The integration of large amounts of public data may allow us to examine general dynamical relationships between regulators and target genes. However, obstacles for such analyses are study-specific biases or batch effects in the original data. Here we present Immuno-Navigator, a batch-corrected gene expression and coexpression database for 24 cell types of the mouse immune system. We systematically removed batch effects from the underlying gene expression data and showed that this removal considerably improved the consistency between inferred correlations and prior knowledge. The data revealed widespread cell type-specific correlation of expression. Integrated analysis tools allow users to use this correlation of expression for the generation of hypotheses about biological networks and candidate regulators in specific cell types. We show several applications of Immuno-Navigator as examples. In one application we successfully predicted known regulators of importance in naturally occurring Treg cells from their expression correlation with a set of Treg-specific genes. For one high-scoring gene, integrin β8 (Itgb8), we confirmed an association between Itgb8 expression in forkhead box P3 (Foxp3)-positive T cells and Treg-specific epigenetic remodeling. Our results also suggest that the regulation of Treg-specific genes within Treg cells is relatively independent of Foxp3 expression, supporting recent results pointing to a Foxp3-independent component in the development of Treg cells.

MIrExpress: A Database for Gene Coexpression Correlation in Immune Cells Based on Mutual Information and Pearson Correlation

Most current gene coexpression databases support the analysis for linear correlation of gene pairs, but not nonlinear correlation of them, which hinders precisely evaluating the gene-gene coexpression strengths. Here, we report a new database, MIrExpress, which takes advantage of the information theory, as well as the Pearson linear correlation method, to measure the linear correlation, nonlinear correlation, and their hybrid of cell-specific gene coexpressions in immune cells. For a given gene pair or probe set pair input by web users, both mutual information (MI) and Pearson correlation coefficient (r) are calculated, and several corresponding values are reported to reflect their coexpression correlation nature, including MI and r values, their respective rank orderings, their rank comparison, and their hybrid correlation value. Furthermore, for a given gene, the top 10 most relevant genes to it are displayed with the MI, r, or their hybrid perspective, respectively. Currently, the database totally includes 16 human cell groups, involving 20,283 human genes. The expression data and the calculated correlation results from the database are interactively accessible on the web page and can be implemented for other related applications and researches.

Progression of mouse skin carcinogenesis is associated with the orchestrated deregulation of mir-200 family members, mir-205 and their common targets

MicroRNAs are small, non-coding RNAs which regulate post-transcriptionally hundreds of target mRNAs. Given that their expression is deregulated in several cancer types, they represent potential diagnostic, prognostic, and predictive biomarkers, as well as next-generation therapeutic targets. Nevertheless, the involvement of miRNAs in non-melanoma skin cancer, a cancer type with increasing prevalence, is not extensively studied, and their comprehensive characterization as regard to the initiation, promotion, and progression stages is missing. To this end, we exploited a well-established multistage mouse skin carcinogenesis model in order to identify miRNAs consistently implicated in different stages of skin carcinogenesis. The cell lines comprising this model were subjected to miRNA expression profiling using microarrays, followed by bioinformatics analysis and validation with Q-PCR, as well as treatment with miRNA modulators. We showed that among all deregulated miRNAs in our system, only a functionally coherent group consisting of the miR-200 family members and miR-205-5p displays a pattern of progressive co-downregulation from the early toward the most aggressive stages of carcinogenesis. Their overlapping, co-regulated putative targets are potentially inter-associated and, of these, the EMT-related Rap1a is overexpressed toward aggressive stages. Ectopic expression of miR-205-5p in spindle cancer cells reduces Rap1a, mitigates cell invasiveness, decreases proliferation, and delays tumor onset. We conclude that deregulation of this miRNA group is primarily associated with aggressive phenotypes of skin cancer cells. Restoration of the miR-205-5p member of this group in spindle cells reduces the expression of critical, co-regulated targets that favor cancer progression, thus reversing the EMT characteristics. © 2015 Wiley Periodicals, Inc.

CellWhere: graphical display of interaction networks organized on subcellular localizations

Given a query list of genes or proteins, CellWhere produces an interactive graphical display that mimics the structure of a cell, showing the local interaction network organized into subcellular locations. This user-friendly tool helps in the formulation of mechanistic hypotheses by enabling the experimental biologist to explore simultaneously two elements of functional context: (i) protein subcellular localization and (ii) protein–protein interactions or gene functional associations. Subcellular localization terms are obtained from public sources (the Gene Ontology and UniProt—together containing several thousand such terms) then mapped onto a smaller number of CellWhere localizations. These localizations include all major cell compartments, but the user may modify the mapping as desired. Protein–protein interaction listings, and their associated evidence strength scores, are obtained from the Mentha interactome server, or power-users may upload a pre-made network produced using some other interactomics tool. The Cytoscape.js JavaScript library is used in producing the graphical display. Importantly, for a protein that has been observed at multiple subcellular locations, users may prioritize the visual display of locations that are of special relevance to their research domain. CellWhere is at http://cellwhere-myology.rhcloud.com.

ImmuCo: a database of gene co-expression in immune cells

Current gene co-expression databases and correlation networks do not support cell-specific analysis. Gene co-expression and expression correlation are subtly different phenomena, although both are likely to be functionally significant. Here, we report a new database, ImmuCo (http://immuco.bjmu.edu.cn), which is a cell-specific database that contains information about gene co-expression in immune cells, identifying co-expression and correlation between any two genes. The strength of co-expression of queried genes is indicated by signal values and detection calls, whereas expression correlation and strength are reflected by Pearson correlation coefficients. A scatter plot of the signal values is provided to directly illustrate the extent of co-expression and correlation. In addition, the database allows the analysis of cell-specific gene expression profile across multiple experimental conditions and can generate a list of genes that are highly correlated with the queried genes. Currently, the database covers 18 human cell groups and 10 mouse cell groups, including 20 283 human genes and 20 963 mouse genes. More than 8.6 × 10⁸ and 7.4 × 10⁸ probe set combinations are provided for querying each human and mouse cell group, respectively. Sample applications support the distinctive advantages of the database.