Differential Network Analysis Research Articles

Abstract P5-04-17: Local network topology differences between early and late recurrence in ER+ breast cancers

Abstract Background: Late recurrence is characteristic of ER+ breast cancers. Despite an apparently effective adjuvant endocrine therapy, many breast cancers recur years after their initial endocrine treatment. Why some tumors recur early (<3 years) and some recur later (>5 years) is poorly understood. If systemic endocrine therapies killed all cells, recurrence would reflect only the appearance of new disease. Thus, we hypothesized that cells that survive and lie dormant may be driven, in part, by altered wiring of their cell death signaling. We, therefore, studied how cell death signaling is differentially wired in primary tumors that will recur early versus those that will recur later. Method: Genes involved in apoptosis, autophagy, ferroptosis, necrosis, and pyroptosis were identified from KEGG to initiate network feature analysis of gene expression data from public and our first in-house gene expression dataset. Data were collected from ER+ breast cancer pre-endocrine treatment samples with up to 20 years follow-up. Publicly available datasets used were GSE6532, GSE2034, GSE7390, GSE17705, GSE12093, and TCGA. We applied our Knowledge-fused Differential Dependency Network (KDDN) analysis tool to the public datasets; KDDN has provided powerful new insights into signaling in breast and other cancers. Common gene-gene interactions (edges) predicted in at least two different datasets were extracted from all KDDN analyses results. To strengthen the relevance of these features, predicted network edges that represent known protein-protein interactions (PPI) were identified from the STRING database, and these edges were noted in the signaling graphs. Final network graphs were constructed using the common edges from all overlaid networks. We conducted IPA analysis on all nodes in the final network and selected those incorporating network hubs. We took a similar approach to our second in-house dataset, which we used for independent testing. Here, patients were included if their tumor exhibited an initial reduction in volume of at least 40% by four months in response to neo-adjuvant Letrozole. Patients were then classified into two groups during follow-up of up to 3.7 years: i) initial tumor size reduction followed by continued response (expected to recur late); ii) initial reduction followed by tumor regrowth (expected to recur early). KDDN analysis was performed on pretreatment samples from these two groups and a network created annotated with PPI information. Results: MAPK8 and CYCS (Molecular Mechanisms of Cancer, p=1.58E-52), TNFRSF1A Neuroinflammation Signaling Pathway, p=1.26E-54), RELA, and NFKB1 (Colorectal Cancer Metastasis Signaling, p=7.94E-35), were identified as hubs. Hubs may be critical signaling components driving the differences between tumors that will become dormant and recur late. Connections between SLC25A6 and SQSTM1 (p = 0.008), BIRC2 and GABARAP (p = 0.021) in the early group, and AKT3 and IRS2 (p = 0.014) in the late group, were shared between the two final networks. With longer follow-up time on the second in-house dataset, we will better define the two groups and identify additional common phenotype specific gene-gene interactions. Citation Format: Clarke R, Dixon M, Jin L, Pearce D, Turnbull A, Selli C, Hu R, Zwart A, Wang Y, Xuan J, Sengupta S, Sims A, Liu MC. Local network topology differences between early and late recurrence in ER+ breast cancers [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-04-17.

DiffNetFDR: differential network analysis with false discovery rate control.

To identify biological network rewiring under different conditions, we develop a user-friendly R package, named DiffNetFDR, to implement two methods developed for testing the difference in different Gaussian graphical models. Compared to existing tools, our methods have the following features: (i) they are based on Gaussian graphical models which can capture the changes of conditional dependencies; (ii) they determine the tuning parameters in a data-driven manner; (iii) they take a multiple testing procedure to control the overall false discovery rate; and (iv) our approach defines the differential network based on partial correlation coefficients so that the spurious differential edges caused by the variants of conditional variances can be excluded. We also develop a Shiny application to provide easier analysis and visualization. Simulation studies are conducted to evaluate the performance of our methods. We also apply our methods to two real gene expression datasets. The effectiveness of our methods is validated by the biological significance of the identified differential networks. R package and Shiny app are available at https://github.com/Zhangxf-ccnu/DiffNetFDR. Supplementary data are available at Bioinformatics online.

The Role of COL5A2 in Patients With Muscle-Invasive Bladder Cancer: A Bioinformatics Analysis of Public Datasets Involving 787 Subjects and 29 Cell Lines.

Bladder cancer (BC) is one of the most common malignancies. Two previous studies identified collagen type V alpha 2 (COL5A2) as a potential biomarker in BC, both are simple reanalysis of a single transcriptomic dataset without subgroup analysis for muscle-invasive BC (MIBC). We focused in MIBC patients and explored the role of COL5A2 from an integration perspective, using refined methodology covering individual participant data meta-analysis and bioinformatics analysis. Eight transcriptomic datasets of 787 MIBC patients (including one dataset containing genomic mutation information) and two drug sensitivity datasets of 29 cell lines in which more than 250 compounds were analyzed. We found subjects with increased COL5A2 gene expression exhibited poorer prognosis, and the power analysis confirmed adequate sample size. FGFR3 was the only gene differential mutated between the COL5A2 high and low expression groups. Differential expression and co-expression network analysis suggested potential association between COL5A2 expression and essential pathways involved in cancer invasion and dissemination, including cell adhesion, extracellular matrix organization, and epithelial-mesenchymal transition. Coordinately, analysis of drug screening datasets and gene-drug interaction also revealed COL5A2 expression linked to cell morphogenesis, angiogenesis, blood vessel development, and urogenital development. The utility and feasibility of COL5A2 for clinically applicable prognosis prediction and risk classification and the exact underlying molecular mechanism should be further investigated in subsequent studies.

Predicting Functional Modules of Liver Cancer Based on Differential Network Analysis.

Complex diseases are generally caused by disorders of biological networks or/and mutations in multiple genes. The efficient and systematic identification of functional modules can not only supply effective diagnosis and treatment in clinic, but also benefit in further in-depth analysis of the pathological mechanism of complex diseases. In this study, we applied the method of differential network to identify functional modules between control and disease samples, which are different from most of the current approaches that focus on differential expression. In particular, we applied our approach to analyze transcriptome data of liver cancer in The Cancer Genome Atlas (TCGA, https://cancergenome.nih.gov/), and we obtained two modules associated with liver cancer. One is a functional gene module that contains a set of liver cancer-related genes, and another is an lncRNA (long non-coding RNA) module that includes liver cancer-related lncRNAs. The results of survival analysis and classification show that the functional modules cannot only be used as effective modular biomarkers to identifying liver cancer, but also predict the prognosis of liver cancer. The method can identify functional modules in genes and lncRNA from liver cancer, and these modules can be used to do prognosis detection and further study in mechanism of liver cancer.

Identification and epigenetic analysis of divergent long non-coding RNAs in multilineage differentiation of human Neural Progenitor Cells

ABSTRACTLong non-coding RNAs have emerged as an important regulatory layer in biological systems. Of the various types of lncRNAs, one class (designated as divergent RNAs/XH), which is in head-to-head overlap with the coding genes, has emerged as a critical biotype that regulates development and cellular differentiation. This work aimed to analyze previously published data on differential expression, epigenetic and network analysis in order to demonstrate the association of divergent lncRNAs, a specific biotype with the differentiation of human neural progenitor cells (hNPCs). We have analyzed various available RNAseq databases that address the neuronal and astrocytic differentiation of hNPCs and identified differentially expressed lncRNAs (DELs) during cell-fate determination. Key DELs identified from the databases were experimentally verified by us in our in-vitro hNPC differentiation system. We also analyzed the change in promoter activity using ChIP-seq datasets of the histone markers H3K4me3 (activation) and H3K27me3 (inactivation) of these DELs. Additionally, we explored the change in the euchromatinization state of DELs (by analyzing DNase-seq data) during lineage-specific differentiation of hNPCs and performed their network analysis. We were able to identify differences between neuronal and astrocytic differentiation of hNPCs at the level of divergent DELs epigenetic markers, DNAase hypersensitive sites and gene expression network. Divergent lncRNAs are more involved in neuronal rather than astrocytic differentiation, while the sense downstream lncRNA biotype appears to be more involved in astrocytic differentiation. By studying the lncRNA involvement of distinct biotypes, we have been able to indicate the preferential role of a particular biotype during lineage-specific differentiation.

Brain transcriptome changes in the aging Drosophila melanogaster accompany olfactory memory performance deficits.

Cognitive decline is a common occurrence of the natural aging process in animals and studying age-related changes in gene expression in the brain might shed light on disrupted molecular pathways that play a role in this decline. The fruit fly is a useful neurobiological model for studying aging due to its short generational time and relatively small brain size. We investigated age-dependent changes in the Drosophila melanogaster whole-brain transcriptome by comparing 5-, 20-, 30- and 40-day-old flies of both sexes. We used RNA-Sequencing of dissected brain samples followed by differential expression, temporal clustering, co-expression network and gene ontology enrichment analyses. We found an overall decline in expression of genes from the mitochondrial oxidative phosphorylation pathway that occurred as part of aging. We also detected, in females, a pattern of continuously declining expression for many neuronal function genes, which was unexpectedly reversed later in life. This group of genes was highly enriched in memory-impairing genes previously identified through an RNAi screen. We also identified deficits in short-term olfactory memory performance in older flies of both sexes, some of which matched the timing of certain changes in the brain transcriptome. Our study provides the first transcriptome profile of aging brains from fruit flies of both sexes, and it will serve as an important resource for those who study aging and cognitive decline in this model.

Comparative Bioinformatics Characteristic of Bladder Cancer Stage 2 from Stage 4 Expression Profile: A Network-Based Study.

Background: Bladder cancer (BC) has remained as one of the most challenging issues in medicine. The aim of this study was to investigate the differential network analysis of stages 2 and 4 of BC to better understand the molecular pathology of these states. Materials and Methods: We chose gene expression data of GSE52519 from Gene Expression Omnibus (GEO) database analyzed by the GEO2R online tool. Cytoscape version 3.6.1 and its algorithms are the methods applied for the network construction and investigation of differentially expressed genes (DEG) in these states. Result: Our result revealed that the analysis DEGs provides useful information about a common molecular feature of stages 2 and 4 of BC. Conclusion: Consequently, the network finding revealed that more investigation about stage 2 is required to achieve an effective therapeutic protocol to block the transition from stage 2 to stage 4.

Comparative Bioinformatics Characteristic of Bladder Cancer Stage 2 from Stage 4 Expression Profile: A Network-Based Study

Background: Bladder cancer (BC) has remained as one of the most challenging issues in medicine. The aim of this study was to investigate the differential network analysis of stages 2 and 4 of BC to better understand the molecular pathology of these states. Materials and Methods: We chose gene expression data of GSE52519 from Gene Expression Omnibus (GEO) database analyzed by the GEO2R online tool. Cytoscape version 3.6.1 and its algorithms are the methods applied for the network construction and investigation of differentially expressed genes (DEG) in these states. Result: Our result revealed that the analysis DEGs provides useful information about a common molecular feature of stages 2 and 4 of BC. Conclusion: Consequently, the network finding revealed that more investigation about stage 2 is required to achieve an effective therapeutic protocol to block the transition from stage 2 to stage 4.[GMJ.2018;7:e1279]

Comparative Bioinformatics Characteristic of Bladder Cancer Stage 2 from Stage 4 Expression Profile: A Network-Based Study

Background: Bladder cancer (BC) has remained as one of the most challenging issues in medicine. The aim of this study was to investigate the differential network analysis of stages 2 and 4 of BC to better understand the molecular pathology of these states. Materials and Methods: We chose gene expression data of GSE52519 from Gene Expression Omnibus (GEO) database analyzed by the GEO2R online tool. Cytoscape version 3.6.1 and its algorithms are the methods applied for the network construction and investigation of differentially expressed genes (DEG) in these states. Result: Our result revealed that the analysis DEGs provides useful information about a common molecular feature of stages 2 and 4 of BC. Conclusion: Consequently, the network finding revealed that more investigation about stage 2 is required to achieve an effective therapeutic protocol to block the transition from stage 2 to stage 4.[GMJ.2018;7:e1279]
 Â

ScdNet: a computational tool for single-cell differential network analysis

BackgroundSingle-cell RNA sequencing (scRNA-Seq) is an emerging technology that has revolutionized the research of the tumor heterogeneity. However, the highly sparse data matrices generated by the technology have posed an obstacle to the analysis of differential gene regulatory networks.ResultsAddressing the challenges, this study presents, as far as we know, the first bioinformatics tool for scRNA-Seq-based differential network analysis (scdNet). The tool features a sample size adjustment of gene-gene correlation, comparison of inter-state correlations, and construction of differential networks. A simulation analysis demonstrated the power of scdNet in the analyses of sparse scRNA-Seq data matrices, with low requirement on the sample size, high computation efficiency, and tolerance of sequencing noises. Applying the tool to analyze two datasets of single circulating tumor cells (CTCs) of prostate cancer and early mouse embryos, our data demonstrated that differential gene regulation plays crucial roles in anti-androgen resistance and early embryonic development.ConclusionsOverall, the tool is widely applicable to datasets generated by the emerging technology to bring biological insights into tumor heterogeneity and other studies. MATLAB implementation of scdNet is available at https://github.com/ChenLabGCCRI/scdNet.

Differential metabolomics analysis allows characterization of diversity of metabolite networks between males and females.

Females and males are known to have different abilities to cope with stress and disease. This study was designed to investigate the effect of sex on properties of a complex interlinked network constructed of central biochemical metabolites. The study involved the blood collection and analysis of a large set of blood metabolic markers from a total of 236 healthy participants, which included 140 females and 96 males. Metabolic profiling yielded concentrations of 168 metabolites for each subject. A differential correlation network analysis approach was developed for this study that allowed detection and characterization of interconnection differences in metabolites in males and females. Through topological analysis of the differential network that depicted metabolite differences in the sexes, we identified metabolites with high centralities in this network. These key metabolites were identified as 10 phosphatidylcholines (PCaaC34:4, PCaaC36:6, PCaaC34:3, PCaaC42:2, PCaeC38:1, PCaeC38:2, PCaaC40:1, PCaeC34:1, PC aa C32:1 and PC aa C40:6) and 4 acylcarnitines (C3-OH, C7-DC, C3 and C0). Identification of these metabolites may help further studies of sex-specific differences in the metabolome that may underlie different responses to stress and disease in males and females.

Predicting and understanding the response to short-term intensive insulin therapy in people with early type 2 diabetes

ObjectiveShort-term intensive insulin therapy (IIT) early in the course of type 2 diabetes acutely improves beta-cell function with long-lasting effects on glycemic control. However, conventional measures cannot determine which patients are better suited for IIT, and little is known about the molecular mechanisms determining response. Therefore, this study aimed to develop a model that could accurately predict the response to IIT and provide insight into molecular mechanisms driving such response in humans. MethodsTwenty-four patients with early type 2 diabetes were assessed at baseline and four weeks after IIT, consisting of basal detemir and premeal insulin aspart. Twelve individuals had a beneficial beta-cell response to IIT (responders) and 12 did not (nonresponders). Beta-cell function was assessed by multiple methods, including Insulin Secretion-Sensitivity Index-2. MicroRNAs (miRNAs) were profiled in plasma samples before and after IIT. The response to IIT was modeled using a machine learning algorithm and potential miRNA-mediated regulatory mechanisms assessed by differential expression, correlation, and functional network analyses (FNA). ResultsBaseline levels of circulating miR-145-5p, miR-29c-3p, and HbA1c accurately (91.7%) predicted the response to IIT (OR = 121 [95% CI: 6.7, 2188.3]). Mechanistically, a previously described regulatory loop between miR-145-5p and miR-483-3p/5p, which controls TP53-mediated apoptosis, appears to also occur in our study population of humans with early type 2 diabetes. In addition, significant (fold change > 2, P < 0.05) longitudinal changes due to IIT in the circulating levels of miR-138-5p, miR-192-5p, miR-195-5p, miR-320b, and let-7a-5p further characterized the responder group and significantly correlated (|r| > 0.4, P < 0.05) with the changes in measures of beta-cell function and insulin sensitivity. FNA identified a network of coordinately/cooperatively regulated miRNA-targeted genes that potentially drives the IIT response through negative regulation of apoptotic processes that underlie beta cell dysfunction and concomitant positive regulation of proliferation. ConclusionsResponses to IIT in people with early type 2 diabetes are associated with characteristic miRNA signatures. This study represents a first step to identify potential responders to IIT (a current limitation in the field) and provides important insight into the pathophysiologic determinants of the reversibility of beta-cell dysfunction.ClinicalTrial.gov identifier: NCT01270789.

Machine learning to identify pairwise interactions between specific IgE antibodies and their association with asthma: A cross-sectional analysis within a population-based birth cohort.

BackgroundThe relationship between allergic sensitisation and asthma is complex; the data about the strength of this association are conflicting. We propose that the discrepancies arise in part because allergic sensitisation may not be a single entity (as considered conventionally) but a collection of several different classes of sensitisation. We hypothesise that pairings between immunoglobulin E (IgE) antibodies to individual allergenic molecules (components), rather than IgE responses to ‘informative’ molecules, are associated with increased risk of asthma.Methods and findingsIn a cross-sectional analysis among 461 children aged 11 years participating in a population-based birth cohort, we measured serum-specific IgE responses to 112 allergen components using a multiplex array (ImmunoCAP Immuno‑Solid phase Allergy Chip [ISAC]). We characterised sensitivity to 44 active components (specific immunoglobulin E [sIgE] > 0.30 units in at least 5% of children) among the 213 (46.2%) participants sensitised to at least one of these 44 components. We adopted several machine learning methodologies that offer a powerful framework to investigate the highly complex sIgE–asthma relationship. Firstly, we applied network analysis and hierarchical clustering (HC) to explore the connectivity structure of component-specific IgEs and identify clusters of component-specific sensitisation (‘component clusters’). Of the 44 components included in the model, 33 grouped in seven clusters (C.sIgE-1–7), and the remaining 11 formed singleton clusters. Cluster membership mapped closely to the structural homology of proteins and/or their biological source. Components in the pathogenesis-related (PR)-10 proteins cluster (C.sIgE-5) were central to the network and mediated connections between components from grass (C.sIgE-4), trees (C.sIgE-6), and profilin clusters (C.sIgE-7) with those in mite (C.sIgE-1), lipocalins (C.sIgE-3), and peanut clusters (C.sIgE-2). We then used HC to identify four common ‘sensitisation clusters’ among study participants: (1) multiple sensitisation (sIgE to multiple components across all seven component clusters and singleton components), (2) predominantly dust mite sensitisation (IgE responses mainly to components from C.sIgE-1), (3) predominantly grass and tree sensitisation (sIgE to multiple components across C.sIgE-4–7), and (4) lower-grade sensitisation. We used a bipartite network to explore the relationship between component clusters, sensitisation clusters, and asthma, and the joint density-based nonparametric differential interaction network analysis and classification (JDINAC) to test whether pairwise interactions of component-specific IgEs are associated with asthma. JDINAC with pairwise interactions provided a good balance between sensitivity (0.84) and specificity (0.87), and outperformed penalised logistic regression with individual sIgE components in predicting asthma, with an area under the curve (AUC) of 0.94, compared with 0.73. We then inferred the differential network of pairwise component-specific IgE interactions, which demonstrated that 18 pairs of components predicted asthma. These findings were confirmed in an independent sample of children aged 8 years who participated in the same birth cohort but did not have component-resolved diagnostics (CRD) data at age 11 years. The main limitation of our study was the exclusion of potentially important allergens caused by both the ISAC chip resolution as well as the filtering step. Clustering and the network analyses might have provided different solutions if additional components had been available.ConclusionsInteractions between pairs of sIgE components are associated with increased risk of asthma and may provide the basis for designing diagnostic tools for asthma.

Large scale study of anti-sense regulation by differential network analysis

BackgroundSystems biology aims to analyse regulation mechanisms into the cell. By mapping interactions observed in different situations, differential network analysis has shown its power to reveal specific cellular responses or specific dysfunctional regulations. In this work, we propose to explore on a large scale the role of natural anti-sense transcription on gene regulation mechanisms, and we focus our study on apple (Malus domestica) in the context of fruit ripening in cold storage.ResultsWe present a differential functional analysis of the sense and anti-sense transcriptomic data that reveals functional terms linked to the ripening process. To develop our differential network analysis, we introduce our inference method of an Extended Core Network; this method is inspired by C3NET, but extends the notion of significant interactions. By comparing two extended core networks, one inferred with sense data and the other one inferred with sense and anti-sense data, our differential analysis is first performed on a local view and reveals AS-impacted genes, genes that have important interactions impacted by anti-sense transcription. The motifs surrounding AS-impacted genes gather transcripts with functions mostly consistent with the biological context of the data used and the method allows us to identify new actors involved in ripening and cold acclimation pathways and to decipher their interactions. Then from a more global view, we compute minimal sub-networks that connect the AS-impacted genes using Steiner trees. Those Steiner trees allow us to study the rewiring of the AS-impacted genes in the network with anti-sense actors.ConclusionAnti-sense transcription is usually ignored in transcriptomic studies. The large-scale differential analysis of apple data that we propose reveals that anti-sense regulation may have an important impact in several cellular stress response mechanisms. Our data mining process enables to highlight specific interactions that deserve further experimental investigations.

TissueEnrich: Tissue-specific gene enrichment analysis.

RNA-Seq data analysis results in lists of genes that may have a similar function, based on differential gene expression analysis or co-expression network analysis. While tools have been developed to identify biological processes that are enriched in the genes sets, there remains a need for tools that identify enrichment of tissue-specific genes. Therefore, we developed TissueEnrich, a tool that calculates tissue-specific gene enrichment in an input gene set. We demonstrated that TissueEnrich can assign tissue identities to single cell clusters and differentiated embryonic stem cells. The TissueEnrich web application is freely available at http://tissueenrich.gdcb.iastate.edu/. The R package is available through Bioconductor at https://bioconductor.org/packages/TissueEnrich. Both the web application and R package are for non-profit academic use under the MIT license. Supplementary data are available at Bioinformatics online.

MetaOmics: analysis pipeline and browser-based software suite for transcriptomic meta-analysis.

The rapid advances of omics technologies have generated abundant genomic data in public repositories and effective analytical approaches are critical to fully decipher biological knowledge inside these data. Meta-analysis combines multiple studies of a related hypothesis to improve statistical power, accuracy and reproducibility beyond individual study analysis. To date, many transcriptomic meta-analysis methods have been developed, yet few thoughtful guidelines exist. Here, we introduce a comprehensive analytical pipeline and browser-based software suite, called MetaOmics, to meta-analyze multiple transcriptomic studies for various biological purposes, including quality control, differential expression analysis, pathway enrichment analysis, differential co-expression network analysis, prediction, clustering and dimension reduction. The pipeline includes many public as well as >10 in-house transcriptomic meta-analytic methods with data-driven and biological-aim-driven strategies, hands-on protocols, an intuitive user interface and step-by-step instructions. MetaOmics is freely available at https://github.com/metaOmics/metaOmics. Supplementary data are available at Bioinformatics online.

DiNAR: revealing hidden patterns of plant signalling dynamics using Differential Network Analysis in R

BackgroundProgress in high-throughput molecular methods accompanied by more complex experimental designs demands novel data visualisation solutions. To specifically answer the question which parts of the specifical biological system are responding in particular perturbation, integrative approach in which experimental data are superimposed on a prior knowledge network is shown to be advantageous.ResultsWe have developed DiNAR, Differential Network Analysis in R, a user-friendly application with dynamic visualisation that integrates multiple condition high-throughput data and extensive biological prior knowledge. Implemented differential network approach and embedded network analysis allow users to analyse condition-specific responses in the context of topology of interest (e.g. immune signalling network) and extract knowledge concerning patterns of signalling dynamics (i.e. rewiring in network structure between two or more biological conditions). We validated the usability of software on the Arabidopsis thaliana and Solanum tuberosum datasets, but it is set to handle any biological instances.ConclusionsDiNAR facilitates detection of network-rewiring events, gene prioritisation for future experimental design and allows capturing dynamics of complex biological system. The fully cross-platform Shiny App is hosted and freely available at https://nib-si.shinyapps.io/DiNAR. The most recent version of the source code is available at https://github.com/NIB-SI/DiNAR/ with a DOI 10.5281/zenodo.1230523 of the archived version in Zenodo.

RNA-seq and flow-cytometry of conventional, scalp, and palmoplantar psoriasis reveal shared and distinct molecular pathways

It has long been recognized that anatomic location is an important feature for defining distinct subtypes of plaque psoriasis. However, little is known about the molecular differences between scalp, palmoplantar, and conventional plaque psoriasis. To investigate the molecular heterogeneity of these psoriasis subtypes, we performed RNA-seq and flow cytometry on skin samples from individuals with scalp, palmoplantar, and conventional plaque psoriasis, along with samples from healthy control patients. We performed differential expression analysis and network analysis using weighted gene coexpression network analysis (WGCNA). Our analysis revealed a core set of 763 differentially expressed genes common to all sub-types of psoriasis. In contrast, we identified 605, 632, and 262 genes uniquely differentially expressed in conventional, scalp, and palmoplantar psoriasis, respectively. WGCNA and pathway analysis revealed biological processes for the core genes as well as subtype-specific genes. Flow cytometry analysis revealed a shared increase in the percentage of CD4+ T regulatory cells in all psoriasis subtypes relative to controls, whereas distinct psoriasis subtypes displayed differences in IL-17A, IFN-gamma, and IL-22 production. This work reveals the molecular heterogeneity of plaque psoriasis and identifies subtype-specific signaling pathways that will aid in the development of therapy that is appropriate for each subtype of plaque psoriasis.

MATHT: A web server for comprehensive transcriptome data analysis

The current software/algorithms for high-throughput sequence data analysis are not user-friendly. We developed MATHT, the Multifaceted Analysis Tool for Human Transcriptome, which is a free web server available at www.biocloudservice.com, to provide more comprehensive and reliable analysis of transcriptome data. The web server provides modules for data preprocessing, differential expression analysis, dataset integration, functional analysis, and network analysis. The sequence and structure analysis module is specially designed for RNA-seq data. MATHT is a user-friendly web server that provides comprehensive analysis of transcriptome data, especially integration analysis using special standardization across different platforms.

RNA-seq assistant: machine learning based methods to identify more transcriptional regulated genes

BackgroundAlthough different quality controls have been applied at different stages of the sample preparation and data analysis to ensure both reproducibility and reliability of RNA-seq results, there are still limitations and bias on the detectability for certain differentially expressed genes (DEGs). Whether the transcriptional dynamics of a gene can be captured accurately depends on experimental design/operation and the following data analysis processes. The workflow of subsequent data processing, such as reads alignment, transcript quantification, normalization, and statistical methods for ultimate identification of DEGs can influence the accuracy and sensitivity of DEGs analysis, producing a certain number of false-positivity or false-negativity. Machine learning (ML) is a multidisciplinary field that employs computer science, artificial intelligence, computational statistics and information theory to construct algorithms that can learn from existing data sets and to make predictions on new data set. ML–based differential network analysis has been applied to predict stress-responsive genes through learning the patterns of 32 expression characteristics of known stress-related genes. In addition, the epigenetic regulation plays critical roles in gene expression, therefore, DNA and histone methylation data has been shown to be powerful for ML-based model for prediction of gene expression in many systems, including lung cancer cells. Therefore, it is promising that ML-based methods could help to identify the DEGs that are not identified by traditional RNA-seq method.ResultsWe identified the top 23 most informative features through assessing the performance of three different feature selection algorithms combined with five different classification methods on training and testing data sets. By comprehensive comparison, we found that the model based on InfoGain feature selection and Logistic Regression classification is powerful for DEGs prediction. Moreover, the power and performance of ML-based prediction was validated by the prediction on ethylene regulated gene expression and the following qRT-PCR.ConclusionsOur study shows that the combination of ML-based method with RNA-seq greatly improves the sensitivity of DEGs identification.