Pathway network inference from gene expression data

PaintOmics 4: new tools for the integrative analysis of multi-omics datasets supported by multiple pathway databases

PaintOmics is a web server for the integrative analysis and visualisation of multi-omics datasets using biological pathway maps. PaintOmics 4 has several notable updates that improve and extend analyses. Three pathway databases are now supported: KEGG, Reactome and MapMan, providing more comprehensive pathway knowledge for animals and plants. New metabolite analysis methods fill gaps in traditional pathway-based enrichment methods. The metabolite hub analysis selects compounds with a high number of significant genes in their neighbouring network, suggesting regulation by gene expression changes. The metabolite class activity analysis tests the hypothesis that a metabolic class has a higher-than-expected proportion of significant elements, indicating that these compounds are regulated in the experiment. Finally, PaintOmics 4 includes a regulatory omics module to analyse the contribution of trans-regulatory layers (microRNA and transcription factors, RNA-binding proteins) to regulate pathways. We show the performance of PaintOmics 4 on both mouse and plant data to highlight how these new analysis features provide novel insights into regulatory biology. PaintOmics 4 is available at https://paintomics.org/.

GAIT-GM integrative cross-omics analyses reveal cholinergic defects in a C. elegans model of Parkinson's disease

Parkinson’s disease (PD) is a disabling neurodegenerative disorder in which multiple cell types, including dopaminergic and cholinergic neurons, are affected. The mechanisms of neurodegeneration in PD are not fully understood, limiting the development of therapies directed at disease-relevant molecular targets. C. elegans is a genetically tractable model system that can be used to disentangle disease mechanisms in complex diseases such as PD. Such mechanisms can be studied combining high-throughput molecular profiling technologies such as transcriptomics and metabolomics. However, the integrative analysis of multi-omics data in order to unravel disease mechanisms is a challenging task without advanced bioinformatics training. Galaxy, a widely-used resource for enabling bioinformatics analysis by the broad scientific community, has poor representation of multi-omics integration pipelines. We present the integrative analysis of gene expression and metabolite levels of a C. elegans PD model using GAIT-GM, a new Galaxy tool for multi-omics data analysis. Using GAIT-GM, we discovered an association between branched-chain amino acid metabolism and cholinergic neurons in the C. elegans PD model. An independent follow-up experiment uncovered cholinergic neurodegeneration in the C. elegans model that is consistent with cholinergic cell loss observed in PD. GAIT-GM is an easy to use Galaxy-based tool for generating novel testable hypotheses of disease mechanisms involving gene-metabolite relationships.

TopoFun: a machine learning method to improve the functional similarity of gene co-expression modules

A comprehensive, accurate functional annotation of genes is key to systems-level approaches. As functionally related genes tend to be co-expressed, one possible approach to identify functional modules or supplement existing gene annotations is to analyse gene co-expression. We describe TopoFun, a machine learning method that combines topological and functional information to improve the functional similarity of gene co-expression modules. Using LASSO, we selected topological descriptors that discriminated modules made of functionally related genes and random modules. Using the selected topological descriptors, we performed linear discriminant analysis to construct a topological score that predicted the type of a module, random-like or functional-like. We combined the topological score with a functional similarity score in a fitness function that we used in a genetic algorithm to explore the co-expression network. To illustrate the use of TopoFun, we started from a subset of the Gene Ontology Biological Processes (GO-BPs) and showed that TopoFun efficiently retrieved genes that we omitted, and aggregated a number of novel genes to the initial GO-BP while improving module topology and functional similarity. Using an independent protein-protein interaction database, we confirmed that the novel genes gathered by TopoFun were functionally related to the original gene set.

The Application of Artificial Intelligence in the Genetic Study of Alzheimer's Disease

Alzheimer's disease (AD) is a neurodegenerative disease in which genetic factors contribute approximately 70% of etiological effects. Studies have found many significant genetic and environmental factors, but the pathogenesis of AD is still unclear. With the application of microarray and next-generation sequencing technologies, research using genetic data has shown explosive growth. In addition to conventional statistical methods for the processing of these data, artificial intelligence (AI) technology shows obvious advantages in analyzing such complex projects. This article first briefly reviews the application of AI technology in medicine and the current status of genetic research in AD. Then, a comprehensive review is focused on the application of AI in the genetic research of AD, including the diagnosis and prognosis of AD based on genetic data, the analysis of genetic variation, gene expression profile, gene-gene interaction in AD, and genetic analysis of AD based on a knowledge base. Although many studies have yielded some meaningful results, they are still in a preliminary stage. The main shortcomings include the limitations of the databases, failing to take advantage of AI to conduct a systematic biology analysis of multilevel databases, and lack of a theoretical framework for the analysis results. Finally, we outlook the direction of future development. It is crucial to develop high quality, comprehensive, large sample size, data sharing resources; a multi-level system biology AI analysis strategy is one of the development directions, and computational creativity may play a role in theory model building, verification, and designing new intervention protocols for AD.

Gene set analysis methods for the functional interpretation of non-mRNA data—Genomic range and ncRNA data

Gene set analysis (GSA) is one of the methods of choice for analyzing the results of current omics studies; however, it has been mainly developed to analyze mRNA (microarray, RNA-Seq) data. The following review includes an update regarding general methods and resources for GSA and then emphasizes GSA methods and tools for non-mRNA omics datasets, specifically genomic range data (ChIP-Seq, SNP and methylation) and ncRNA data (miRNAs, lncRNAs and others). In the end, the state of the GSA field for non-mRNA datasets is discussed, and some current challenges and trends are highlighted, especially the use of network approaches to face complexity issues.

PaintOmics 3: a web resource for the pathway analysis and visualization of multi-omics data

The increasing availability of multi-omic platforms poses new challenges to data analysis. Joint visualization of multi-omics data is instrumental in better understanding interconnections across molecular layers and in fully utilizing the multi-omic resources available to make biological discoveries. We present here PaintOmics 3, a web-based resource for the integrated visualization of multiple omic data types onto KEGG pathway diagrams. PaintOmics 3 combines server-end capabilities for data analysis with the potential of modern web resources for data visualization, providing researchers with a powerful framework for interactive exploration of their multi-omics information. Unlike other visualization tools, PaintOmics 3 covers a comprehensive pathway analysis workflow, including automatic feature name/identifier conversion, multi-layered feature matching, pathway enrichment, network analysis, interactive heatmaps, trend charts, and more. It accepts a wide variety of omic types, including transcriptomics, proteomics and metabolomics, as well as region-based approaches such as ATAC-seq or ChIP-seq data. The tool is freely available at www.paintomics.org.

Extreme learning machines for reverse engineering of gene regulatory networks from expression time series

Motivation

The reconstruction of gene regulatory networks (GRNs) from genes profiles has a growing interest in bioinformatics for understanding the complex regulatory mechanisms in cellular systems. GRNs explicitly represent the cause-effect of regulation among a group of genes and its reconstruction is today a challenging computational problem. Several methods were proposed, but most of them require different input sources to provide an acceptable prediction. Thus, it is a great challenge to reconstruct a GRN only from temporal gene expression data.

Results

Extreme Learning Machine (ELM) is a new supervised neural model that has gained interest in the last years because of its higher learning rate and better performance than existing supervised models in terms of predictive power. This work proposes a novel approach for GRNs reconstruction in which ELMs are used for modeling the relationships between gene expression time series. Artificial datasets generated with the well-known benchmark tool used in DREAM competitions were used. Real datasets were used for validation of this novel proposal with well-known GRNs underlying the time series. The impact of increasing the size of GRNs was analyzed in detail for the compared methods. The results obtained confirm the superiority of the ELM approach against very recent state-of-the-art methods in the same experimental conditions.

Availability and implementation

The web demo can be found at http://sinc.unl.edu.ar/web-demo/elm-grnnminer/. The source code is available at https://sourceforge.net/projects/sourcesinc/files/elm-grnnminer.

Contact

mrubiolo@santafe-conicet.gov.ar.

Supplementary information

Supplementary data are available at Bioinformatics online.

Different responses of colorectal cancer cells to alternative sequences of cetuximab and oxaliplatin

Therapeutic protocols including EGFR antibodies in the context of oxaliplatin-based regimens have variable clinical effect in colorectal cancer. Here, we tested the effect of the EGFR antibody cetuximab in different sequential combinations with oxaliplatin on the growth of colorectal cancer cells in vitro and in vivo. Cetuximab reduced the efficacy of oxaliplatin when administered before oxaliplatin but provided additive effect when administered after oxaliplatin regardless of the KRAS or BRAF mutation status of the cells. Systemic gene expression and protein phosphorylation screens revealed alternatively activated pathways regulating apoptosis, cell cycle and DNA damage response. Functional assays indicated that cetuximab-induced arrest of the cells into the G1 phase of the cell cycle was associated with reduced responsiveness of the cells to subsequent treatment with oxaliplatin. In contrast, oxaliplatin-enhanced responsiveness to subsequent treatment with cetuximab was associated with increased apoptosis, inhibition of STAT3 activity and increased EGFR down-regulation. This preclinical study indicates that optimizing the sequence of administration may enhance the antitumor effect of combination therapy with EGFR antibodies and oxaliplatin.

Analysis of Gene Expression Variance in Schizophrenia Using Structural Equation Modeling

Schizophrenia (SCZ) is a psychiatric disorder of unknown etiology. There is evidence suggesting that aberrations in neurodevelopment are a significant attribute of schizophrenia pathogenesis and progression. To identify biologically relevant molecular abnormalities affecting neurodevelopment in SCZ we used cultured neural progenitor cells derived from olfactory neuroepithelium (CNON cells). Here, we tested the hypothesis that variance in gene expression differs between individuals from SCZ and control groups. In CNON cells, variance in gene expression was significantly higher in SCZ samples in comparison with control samples. Variance in gene expression was enriched in five molecular pathways: serine biosynthesis, PI3K-Akt, MAPK, neurotrophin and focal adhesion. More than 14% of variance in disease status was explained within the logistic regression model (C-value = 0.70) by predictors accounting for gene expression in 69 genes from these five pathways. Structural equation modeling (SEM) was applied to explore how the structure of these five pathways was altered between SCZ patients and controls. Four out of five pathways showed differences in the estimated relationships among genes: between KRAS and NF1, and KRAS and SOS1 in the MAPK pathway; between PSPH and SHMT2 in serine biosynthesis; between AKT3 and TSC2 in the PI3K-Akt signaling pathway; and between CRK and RAPGEF1 in the focal adhesion pathway. Our analysis provides evidence that variance in gene expression is an important characteristic of SCZ, and SEM is a promising method for uncovering altered relationships between specific genes thus suggesting affected gene regulation associated with the disease. We identified altered gene-gene interactions in pathways enriched for genes with increased variance in expression in SCZ. These pathways and loci were previously implicated in SCZ, providing further support for the hypothesis that gene expression variance plays important role in the etiology of SCZ.

Identification and characterization of some putative genes involved in arabinoxylan biosynthesis in Plantago ovata

Plantago ovata is an important source of Psyllium (Isabgol), which swells upon contact with water forming mucilaginous mass, largely composed of arabinoxylans. In this study, we analyzed the expression pattern of arabinoxylan biosynthetic pathway genes at different stages of seed development in P. ovata. Besides, arabinoxylans were quantified at different stages of seed development in water extractable and water unextractable fractions. The expression analysis revealed 5-8 fold increase in the levels of expression of some genes involved in arabinoxylan biosynthetic pathway such as UDP-arabinopyranose mutase, UDP-xylosyltransferase 2 and xylan glucuronosyltransferase at 15 days after pollination stage in seed. The xylose and arabinose units were analyzed at different stages of seed development and also in water-soluble (cold water and hot water), alkali and ethanolic fractions. The concentration of xylose and arabinose units increased steadily after pollination. Overall, alkali extract had high concentration of xylose (0.70 ± 0.022 mg/g) and arabinose units (0.10 ± 0.01 mg/g) at 15 days after pollination stage.

Analyzing the genes related to Alzheimer's disease via a network and pathway-based approach

Background

Our understanding of the molecular mechanisms underlying Alzheimer’s disease (AD) remains incomplete. Previous studies have revealed that genetic factors provide a significant contribution to the pathogenesis and development of AD. In the past years, numerous genes implicated in this disease have been identified via genetic association studies on candidate genes or at the genome-wide level. However, in many cases, the roles of these genes and their interactions in AD are still unclear. A comprehensive and systematic analysis focusing on the biological function and interactions of these genes in the context of AD will therefore provide valuable insights to understand the molecular features of the disease.

Method

In this study, we collected genes potentially associated with AD by screening publications on genetic association studies deposited in PubMed. The major biological themes linked with these genes were then revealed by function and biochemical pathway enrichment analysis, and the relation between the pathways was explored by pathway crosstalk analysis. Furthermore, the network features of these AD-related genes were analyzed in the context of human interactome and an AD-specific network was inferred using the Steiner minimal tree algorithm.

Results

We compiled 430 human genes reported to be associated with AD from 823 publications. Biological theme analysis indicated that the biological processes and biochemical pathways related to neurodevelopment, metabolism, cell growth and/or survival, and immunology were enriched in these genes. Pathway crosstalk analysis then revealed that the significantly enriched pathways could be grouped into three interlinked modules—neuronal and metabolic module, cell growth/survival and neuroendocrine pathway module, and immune response-related module—indicating an AD-specific immune-endocrine-neuronal regulatory network. Furthermore, an AD-specific protein network was inferred and novel genes potentially associated with AD were identified.

Conclusion

By means of network and pathway-based methodology, we explored the pathogenetic mechanism underlying AD at a systems biology level. Results from our work could provide valuable clues for understanding the molecular mechanism underlying AD. In addition, the framework proposed in this study could be used to investigate the pathological molecular network and genes relevant to other complex diseases or phenotypes.

Electronic supplementary material

The online version of this article (doi:10.1186/s13195-017-0252-z) contains supplementary material, which is available to authorized users.

Revealing complex function, process and pathway interactions with high-throughput expression and biological annotation data

The biological relationships both between and within the functions, processes and pathways that operate within complex biological systems are only poorly characterized, making the interpretation of large scale gene expression datasets extremely challenging. Here, we present an approach that integrates gene expression and biological annotation data to identify and describe the interactions between biological functions, processes and pathways that govern a phenotype of interest. The product is a global, interconnected network, not of genes but of functions, processes and pathways, that represents the biological relationships within the system. We validated our approach on two high-throughput expression datasets describing organismal and organ development. Our findings are well supported by the available literature, confirming that developmental processes and apoptosis play key roles in cell differentiation. Furthermore, our results suggest that processes related to pluripotency and lineage commitment, which are known to be critical for development, interact mainly indirectly, through genes implicated in more general biological processes. Moreover, we provide evidence that supports the relevance of cell spatial organization in the developing liver for proper liver function. Our strategy can be viewed as an abstraction that is useful to interpret high-throughput data and devise further experiments.

Subpathway-CorSP: Identification of metabolic subpathways via integrating expression correlations and topological features between metabolites and genes of interest within pathways

Metabolic pathway analysis is a popular strategy for comprehensively researching metabolites and genes of interest associated with specific diseases. However, the traditional pathway identification methods do not accurately consider the combined effect of these interesting molecules and neglects expression correlations or topological features embedded in the pathways. In this study, we propose a powerful method, Subpathway-CorSP, for identifying metabolic subpathway regions. This method improved on original pathway identification methods by using a subpathway identification strategy and emphasizing expression correlations between metabolites and genes of interest based on topological features within the metabolic pathways. We analyzed a prostate cancer data set and its metastatic sub-group data set with detailed comparison of Subpathway-CorSP with four traditional pathway identification methods. Subpathway-CorSP was able to identify multiple subpathway regions whose entire corresponding pathways were not detected by traditional pathway identification methods. Further evidences indicated that Subpathway-CorSP provided a robust and efficient way of reliably recalling cancer-related subpathways and locating novel subpathways by the combined effect of metabolites and genes. This was a novel subpathway strategy based on systematically considering expression correlations and topological features between metabolites and genes of interest within given pathways.

Multi-omic data integration and analysis using systems genomics approaches: methods and applications in animal production, health and welfare

In the past years, there has been a remarkable development of high-throughput omics (HTO) technologies such as genomics, epigenomics, transcriptomics, proteomics and metabolomics across all facets of biology. This has spearheaded the progress of the systems biology era, including applications on animal production and health traits. However, notwithstanding these new HTO technologies, there remains an emerging challenge in data analysis. On the one hand, different HTO technologies judged on their own merit are appropriate for the identification of disease-causing genes, biomarkers for prevention and drug targets for the treatment of diseases and for individualized genomic predictions of performance or disease risks. On the other hand, integration of multi-omic data and joint modelling and analyses are very powerful and accurate to understand the systems biology of healthy and sustainable production of animals. We present an overview of current and emerging HTO technologies each with a focus on their applications in animal and veterinary sciences before introducing an integrative systems genomics framework for analysing and integrating multi-omic data towards improved animal production, health and welfare. We conclude that there are big challenges in multi-omic data integration, modelling and systems-level analyses, particularly with the fast emerging HTO technologies. We highlight existing and emerging systems genomics approaches and discuss how they contribute to our understanding of the biology of complex traits or diseases and holistic improvement of production performance, disease resistance and welfare.

An overview of bioinformatics methods for modeling biological pathways in yeast

The advent of high-throughput genomics techniques, along with the completion of genome sequencing projects, identification of protein-protein interactions and reconstruction of genome-scale pathways, has accelerated the development of systems biology research in the yeast organism Saccharomyces cerevisiae In particular, discovery of biological pathways in yeast has become an important forefront in systems biology, which aims to understand the interactions among molecules within a cell leading to certain cellular processes in response to a specific environment. While the existing theoretical and experimental approaches enable the investigation of well-known pathways involved in metabolism, gene regulation and signal transduction, bioinformatics methods offer new insights into computational modeling of biological pathways. A wide range of computational approaches has been proposed in the past for reconstructing biological pathways from high-throughput datasets. Here we review selected bioinformatics approaches for modeling biological pathways inS. cerevisiae, including metabolic pathways, gene-regulatory pathways and signaling pathways. We start with reviewing the research on biological pathways followed by discussing key biological databases. In addition, several representative computational approaches for modeling biological pathways in yeast are discussed.

Leveraging global gene expression patterns to predict expression of unmeasured genes

Background

Large collections of paraffin-embedded tissue represent a rich resource to test hypotheses based on gene expression patterns; however, measurement of genome-wide expression is cost-prohibitive on a large scale. Using the known expression correlation structure within a given disease type (in this case, high grade serous ovarian cancer; HGSC), we sought to identify reduced sets of directly measured (DM) genes which could accurately predict the expression of a maximized number of unmeasured genes.

Results

We developed a greedy gene set selection (GGS) algorithm which returns a DM set of user specified size based on a specific correlation threshold (|r_P|) and minimum number of DM genes that must be correlated to an unmeasured gene in order to infer the value of the unmeasured gene (redundancy). We evaluated GGS in the Cancer Genome Atlas (TCGA) HGSC data across 144 combinations of DM size, redundancy (1–3), and |r_P| (0.60, 0.65, 0.70). Across the parameter sweep, GGS allows on average 9 times more gene expression information to be captured compared to the DM set alone. GGS successfully augments prognostic HGSC gene sets; the addition of 20 GGS selected genes more than doubles the number of genes whose expression is predictable. Moreover, the expression prediction is highly accurate. After training regression models for the predictable gene set using 2/3 of the TCGA data, the average accuracy (ranked correlation of true and predicted values) in the 1/3 testing partition and four independent populations is above 0.65 and approaches 0.8 for conservative parameter sets. We observe similar accuracies in the TCGA HGSC RNA-sequencing data. Specifically, the prediction accuracy increases with increasing redundancy and increasing |r_P|.

Conclusions

GGS-selected genes, which maximize expression information about unmeasured genes, can be combined with candidate gene sets as a cost effective way to increase the amount of gene expression information obtained in large studies. This method can be applied to any organism, model system, disease, or tissue type for which whole genome gene expression data exists.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-2250-5) contains supplementary material, which is available to authorized users.

Insights into protein interaction networks reveal non-receptor kinases as significant druggable targets for psoriasis

Psoriasis is a chronic disease of the skin characterized by hyper proliferation and inflammation of the epidermis and dermal components of the skin. T-cell-dependent inflammatory process in skin governs the pathogenesis of psoriasis. An in-silico search strategy was utilized to identify psoriatic therapeutic drug targets. The gene expression profiling of psoriatic skin identified a total of 427 differentially expressed genes (DEGs). Gene ontology investigation of DEGs identified genes involved in calcium binding, apoptosis, keratinisation, lipid transportation and homeostasis apart from immune mediated processes. The protein interaction networks identified proteins involved in various signaling mechanisms with high degree of interconnections. The gene modules derived from the main network were enriched with rich kinome. These sub-networks were dominated by the presence of non-receptor kinase family members which are major signal transmitters in immune response. The computational approach has aided in the identification of non-receptor kinases as potential targets for psoriasis drug development.

Detecting Cancer Pathway Crosstalk with Distance Correlation

Biological pathway regulation is complex, yet it underlies the functional coordination in a cell. Cancer is a disease that is characterized by unregulated growth, driven by underlying pathway deregulation. This pathway deregulation is both within pathways and between pathways. Here, we propose a method to detect inter-pathway coordination using distance correlation. Utilizing data generated from microarray experiments, we separate the genes into pathways and calculate the pairwise distance correlation between them. The result is intuitively viewed as a network of differentially dependent pathways. We find intuitive, yet surprising significant hub pathways, including glycophosphatidylinositol anchor synthesis in lung cancer.

Gene network biological validity based on gene-gene interaction relevance

In recent years, gene networks have become one of the most useful tools for modeling biological processes. Many inference gene network algorithms have been developed as techniques for extracting knowledge from gene expression data. Ensuring the reliability of the inferred gene relationships is a crucial task in any study in order to prove that the algorithms used are precise. Usually, this validation process can be carried out using prior biological knowledge. The metabolic pathways stored in KEGG are one of the most widely used knowledgeable sources for analyzing relationships between genes. This paper introduces a new methodology, GeneNetVal, to assess the biological validity of gene networks based on the relevance of the gene-gene interactions stored in KEGG metabolic pathways. Hence, a complete KEGG pathway conversion into a gene association network and a new matching distance based on gene-gene interaction relevance are proposed. The performance of GeneNetVal was established with three different experiments. Firstly, our proposal is tested in a comparative ROC analysis. Secondly, a randomness study is presented to show the behavior of GeneNetVal when the noise is increased in the input network. Finally, the ability of GeneNetVal to detect biological functionality of the network is shown.

Data integration in the era of omics: current and future challenges

To integrate heterogeneous and large omics data constitutes not only a conceptual challenge but a practical hurdle in the daily analysis of omics data. With the rise of novel omics technologies and through large-scale consortia projects, biological systems are being further investigated at an unprecedented scale generating heterogeneous and often large data sets. These data-sets encourage researchers to develop novel data integration methodologies. In this introduction we review the definition and characterize current efforts on data integration in the life sciences. We have used a web-survey to assess current research projects on data-integration to tap into the views, needs and challenges as currently perceived by parts of the research community.