SLEPR: a sample-level enrichment-based pathway ranking method -- seeking biological themes through pathway-level consistency

CeRNA network analysis and functional enrichment of salt sensitivity of blood pressure by weighted-gene co-expression analysis

Background

Salt sensitivity of blood pressure (SSBP) is an independent risk factor for cardiovascular disease. The pathogenic mechanisms of SSBP are still uncertain. This study aimed to construct the co-regulatory network of SSBP and data mining strategy based on the competitive endogenous RNA (ceRNA) theory.

Methods

LncRNA and mRNA microarray was performed to screen for candidate RNAs. Four criteria were used to select the potential differently expressed RNAs. The weighted correlation network analysis (WGCNA) package of R software and target miRNA and mRNA prediction online databases were used to construct the ceRNA co-regulatory network and discover the pathways related to SSBP. Gene ontology enrichment, gene set enrichment analysis (GSEA) and KEGG pathway analysis were performed to explore the functions of hub genes in networks.

Results

There were 274 lncRNAs and 36 mRNAs that differently expressed between salt-sensitive and salt-resistant groups (P < 0.05). Using WGCNA analysis, two modules were identified (blue and turquoise). The blue module had a positive relationship with salt-sensitivity (R = 0.7, P < 0.01), high-density lipoprotein (HDL) (R = 0.53, P = 0.02), and total cholesterol (TC) (R = 0.55, P = 0.01). The turquoise module was positively related with triglyceride (TG) (R = 0.8, P < 0.01) and low-density lipoprotein (LDL) (R = 0.54, P = 0.01). Furthermore, 84 ceRNA loops were identified and one loop may be of great importance for involving in pathogenesis of SSBP. KEGG analysis showed that differently expressed mRNAs were mostly enriched in the SSBP-related pathways. However, the enrichment results of GSEA were mainly focused on basic physical metabolic processes.

Conclusion

The microarray data mining process based on WGCNA co-expression analysis had identified 84 ceRNA loops that closely related with known SSBP pathogenesis. The results of our study provide implications for further understanding of the pathogenesis of SSBP and facilitate the precise diagnosis and therapeutics.

IFNL4-ΔG Allele Is Associated with an Interferon Signature in Tumors and Survival of African-American Men with Prostate Cancer

Purpose: Men of African ancestry experience an excessive prostate cancer mortality that could be related to an aggressive tumor biology. We previously described an immune-inflammation signature in prostate tumors of African-American (AA) patients. Here, we further deconstructed this signature and investigated its relationships with tumor biology, survival, and a common germline variant in the IFNλ4 (IFNL4) gene.Experimental Design: We analyzed gene expression in prostate tissue datasets and performed genotype and survival analyses. We also overexpressed IFNL4 in human prostate cancer cells.Results: We found that a distinct interferon (IFN) signature that is analogous to the previously described "IFN-related DNA damage resistance signature" (IRDS) occurs in prostate tumors. Evaluation of two independent patient cohorts revealed that IRDS is detected about twice as often in prostate tumors of AA than European-American men. Furthermore, analysis in TCGA showed an association of increased IRDS in prostate tumors with decreased disease-free survival. To explain these observations, we assessed whether IRDS is associated with an IFNL4 germline variant (rs368234815-ΔG) that controls production of IFNλ4, a type III IFN, and is most common in individuals of African ancestry. We show that the IFNL4 rs368234815-ΔG allele was significantly associated with IRDS in prostate tumors and overall survival of AA patients. Moreover, IFNL4 overexpression induced IRDS in three human prostate cancer cell lines.Conclusions: Our study links a germline variant that controls production of IFNλ4 to the occurrence of a clinically relevant IFN signature in prostate tumors that may predominantly affect men of African ancestry. Clin Cancer Res; 24(21); 5471-81. ©2018 AACR.

A computational method using differential gene expression to predict altered metabolism of multicellular organisms

Altered metabolism is often identified as a cause or an effect of physiology and pathogenesis. But it is difficult to predict the metabolic flux distributions of multicellular organisms due to the lack of an explicit metabolic objective function. Here we present a computational method which can successfully describe the differences in metabolism between two different conditions on a large scale. By integrating gene expression data with an existing comprehensive reconstruction of the global human metabolic network, we qualitatively predicted significantly differential fluxes without prior knowledge or the rate of metabolite uptake and secretion. Therefore, this method can be applied for both microorganisms and multicellular organisms. Different from traditional enrichment analysis methods and constraint-based models, we consider conditions and interactions within the metabolic network simultaneously. To apply the proposed method, we predicted altered fluxes for E. coli strains and clear cell renal cell carcinoma, while the E. coli strains are growing aerobically in a chemostat with different dilution rates and clear cell renal cell carcinoma is compared with normal kidney cells. Then we map the significantly differential reactions to metabolic subsystems defined in the original metabolic network for ccRCC to observe the altered metabolism. In contrast with existing studies, our results show a high accuracy of the E. coli experiment and a more reasonable prediction of the ccRCC experiment. The method presented here provides a computational approach for the genome-wide study of altered metabolism under pairs of conditions for both microorganisms and multicellular organisms.

Integrative enrichment analysis: a new computational method to detect dysregulated pathways in heterogeneous samples

BACKGROUND

Pathway enrichment analysis is a useful tool to study biology and biomedicine, due to its functional screening on well-defined biological procedures rather than separate molecules. The measurement of malfunctions of pathways with a phenotype change, e.g., from normal to diseased, is the key issue when applying enrichment analysis on a pathway. The differentially expressed genes (DEGs) are widely focused in conventional analysis, which is based on the great purity of samples. However, the disease samples are usually heterogeneous, so that, the genes with great differential expression variance (DEVGs) are becoming attractive and important to indicate the specific state of a biological system. In the context of differential expression variance, it is still a challenge to measure the enrichment or status of a pathway. To address this issue, we proposed Integrative Enrichment Analysis (IEA) based on a novel enrichment measurement.

RESULTS

The main competitive ability of IEA is to identify dysregulated pathways containing DEGs and DEVGs simultaneously, which are usually under-scored by other methods. Next, IEA provides two additional assistant approaches to investigate such dysregulated pathways. One is to infer the association among identified dysregulated pathways and expected target pathways by estimating pathway crosstalks. The other one is to recognize subtype-factors as dysregulated pathways associated to particular clinical indices according to the DEVGs' relative expressions rather than conventional raw expressions. Based on a previously established evaluation scheme, we found that, in particular cohorts (i.e., a group of real gene expression datasets from human patients), a few target disease pathways can be significantly high-ranked by IEA, which is more effective than other state-of-the-art methods. Furthermore, we present a proof-of-concept study on Diabetes to indicate: IEA rather than conventional ORA or GSEA can capture the under-estimated dysregulated pathways full of DEVGs and DEGs; these newly identified pathways could be significantly linked to prior-known disease pathways by estimated crosstalks; and many candidate subtype-factors recognized by IEA also have significant relation with the risk of subtypes of genotype-phenotype associations.

CONCLUSIONS

Totally, IEA supplies a new tool to carry on enrichment analysis in the complicate context of clinical application (i.e., heterogeneity of disease), as a necessary complementary and cooperative approach to conventional ones.

ChiNet uncovers rewired transcription subnetworks in tolerant yeast for advanced biofuels conversion

Analysis of rewired upstream subnetworks impacting downstream differential gene expression aids the delineation of evolving molecular mechanisms. Cumulative statistics based on conventional differential correlation are limited for subnetwork rewiring analysis since rewiring is not necessarily equivalent to change in correlation coefficients. Here we present a computational method ChiNet to quantify subnetwork rewiring by statistical heterogeneity that enables detection of potential genotype changes causing altered transcription regulation in evolving organisms. Given a differentially expressed downstream gene set, ChiNet backtracks a rewired upstream subnetwork from a super-network including gene interactions known to occur under various molecular contexts. We benchmarked ChiNet for its high accuracy in distinguishing rewired artificial subnetworks, in silico yeast transcription-metabolic subnetworks, and rewired transcription subnetworks for Candida albicans versus Saccharomyces cerevisiae, against two differential-correlation based subnetwork rewiring approaches. Then, using transcriptome data from tolerant S. cerevisiae strain NRRL Y-50049 and a wild-type intolerant strain, ChiNet identified 44 metabolic pathways affected by rewired transcription subnetworks anchored to major adaptively activated transcription factor genes YAP1, RPN4, SFP1 and ROX1, in response to toxic chemical challenges involved in lignocellulose-to-biofuels conversion. These findings support the use of ChiNet in rewiring analysis of subnetworks where differential interaction patterns resulting from divergent nonlinear dynamics abound.

Systems analysis utilising pathway interactions identifies sonic hedgehog pathway as a primary biomarker and oncogenic target in hepatocellular carcinoma

The development and progression of cancer is associated with disruption of biological networks. Historically studies have identified sets of signature genes involved in events ultimately leading to the development of cancer. Identification of such sets does not indicate which biologic processes are oncogenic drivers and makes it difficult to identify key networks to target for interventions. Using a comprehensive, integrated computational approach, the authors identify the sonic hedgehog (SHH) pathway as the gene network that most significantly distinguishes tumour and tumour-adjacent samples in human hepatocellular carcinoma (HCC). The analysis reveals that the SHH pathway is commonly activated in the tumour samples and its activity most significantly differentiates tumour from the non-tumour samples. The authors experimentally validate these in silico findings in the same biologic material using Western blot analysis. This analysis reveals that the expression levels of SHH, phosphorylated cyclin B1, and CDK7 levels are much higher in most tumour tissues as compared to normal tissue. It is also shown that siRNA-mediated silencing of SHH gene expression resulted in a significant reduction of cell proliferation in a liver cancer cell line, SNU449 indicating that SHH plays a major role in promoting cell proliferation in liver cancer. The SHH pathway is a key network underpinning HCC aetiology which may guide the development of interventions for this most common form of human liver cancer.

Analysis considerations for utilizing RNA-Seq to characterize the brain transcriptome

RNA-Seq allows one to examine only gene expression as well as expression of noncoding RNAs, alternative splicing, and allele-specific expression. With this increased sensitivity and dynamic range, there are computational and statistical considerations that need to be contemplated, which are highly dependent on the biological question being asked. We highlight these to provide an overview of their importance and the impact they can have on downstream interpretation of the brain transcriptome.

Comparison of methods to identify aberrant expression patterns in individual patients: augmenting our toolkit for precision medicine

BACKGROUND

Patient-specific aberrant expression patterns in conjunction with functional screening assays can guide elucidation of the cancer genome architecture and identification of therapeutic targets. Since most statistical methods for expression analysis are focused on differences between experimental groups, the performance of approaches for patient-specific expression analyses are currently less well characterized. A comparison of methods for the identification of genes that are dysregulated relative to a single sample in a given set of experimental samples, to our knowledge, has not been performed.

METHODS

We systematically evaluated several methods including variations on the nearest neighbor based outlying degree method, as well as the Zscore and a robust variant for their suitability to detect patient-specific events. The methods were assessed using both simulations and expression data from a cohort of pediatric acute B lymphoblastic leukemia patients.

RESULTS

We first assessed power and false discovery rates using simulations and found that even under optimal conditions, high effect sizes (>4 unit differences) were necessary to have acceptable power for any method (>0.9) though high false discovery rates (>0.1) were pervasive across simulation conditions. Next we introduced a technical factor into the simulation and found that performance was reduced for all methods and that using weights with the outlying degree could provide performance gains depending on the number of samples and genes affected by the technical factor. In our use case that highlights the integration of functional assays and aberrant expression in a patient cohort (the identification of gene dysregulation events associated with the targets from a siRNA screen), we demonstrated that both the outlying degree and the Zscore can successfully identify genes dysregulated in one patient sample. However, only the outlying degree can identify genes dysregulated across several patient samples.

CONCLUSION

Our results show that outlying degree methods may be a useful alternative to the Zscore or Rscore in a personalized medicine context especially in small to medium sized (between 10 and 50 samples) expression datasets with moderate to high sample-to-sample variability. From these results we provide guidelines for detection of aberrant expression in a precision medicine context.

Silencing of miR-148a in cancer-associated fibroblasts results in WNT10B-mediated stimulation of tumor cell motility

The tumor microenvironment has an important role in cancer progression. Here we show that miR-148a is downregulated in 15 out of 16 samples (94%) of cancer-associated fibroblasts (CAFs) compared with matched normal tissue fibroblasts (NFs) established from patients with endometrial cancer. Laser-capture microdissection of stromal cells from normal tissue and endometrial cancer confirmed this observation. Treatment of cells with 5-aza-deoxycytidine stimulated the expression of miR-148a in the majority of CAFs implicating DNA methylation in the regulation of miR-148a expression. Investigation of miR-148a function in fibroblasts demonstrated that conditioned media (CM) from CAFs overexpressing miR-148a significantly impaired the migration of five endometrial cancer cell lines without affecting their growth rates in co-culture experiments. Among predicted miR-148a target genes are two WNT family members, WNT1 and WNT10B. Activation of the WNT/β-catenin pathway in CAFs was confirmed by microarray analysis of gene expression and increased activity of the SuperTOPFlash luciferase reporter. We found elevated levels of WNT10B protein in CAFs and its level decreased when miR-148a was re-introduced by lentiviral infection. The 3'-UTR of WNT10B, cloned downstream of luciferase cDNA, suppressed luciferase activity when co-expressed with miR-148a indicating that WNT10B is a direct target of miR-148a. In contrast to the effect of miR-148a, WNT10B stimulated migration of endometrial cancer cell lines. Our findings have defined a molecular mechanism in the tumor microenvironment that is a novel target for cancer therapy.

Perturbation of Rb, p53, and Brca1 or Brca2 cooperate in inducing metastatic serous epithelial ovarian cancer

The majority of human high-grade serous epithelial ovarian cancer (SEOC) is characterized by frequent mutations in p53 and alterations in the RB and FOXM1 pathways. A subset of human SEOC harbors a combination of germline and somatic mutations as well as epigenetic dysfunction for BRCA1/2. Using Cre-conditional alleles and intrabursal induction by Cre-expressing adenovirus in genetically engineered mice, we analyzed the roles of pathway perturbations in epithelial ovarian cancer initiation and progression. Inactivation of RB-mediated tumor suppression induced surface epithelial proliferation with progression to stage I carcinoma. Additional biallelic inactivation and/or missense p53 mutation in the presence or absence of Brca1/2 caused progression to stage IV disease. As in human SEOC, mice developed peritoneal carcinomatosis, ascites, and distant metastases. Unbiased gene expression and metabolomic profiling confirmed that Rb, p53, and Brca1/2-triple mutant tumors aligned with human SEOC, and not with other intraperitoneal cancers. Together, our findings provide a novel resource for evaluating disease etiology and biomarkers, therapeutic evaluation, and improved imaging strategies in epithelial ovarian cancer.

Sample-level enrichment analysis unravels shared stress phenotypes among multiple cancer types

BACKGROUND

Adaptation to stress signals in the tumor microenvironment is a crucial step towards carcinogenic phenotype. The adaptive alterations attained by cells to withstand different types of insults are collectively referred to as the stress phenotypes of cancers. In this manuscript we explore the interrelation of different stress phenotypes in multiple cancer types and ask if these phenotypes could be used to explain prognostic differences among tumor samples.

METHODS

We propose a new approach based on enrichment analysis at the level of samples (sample-level enrichment analysis - SLEA) in expression profiling datasets. Without using a priori phenotypic information about samples, SLEA calculates an enrichment score per sample per gene set using z-test. This score is used to determine the relative importance of the corresponding pathway or module in different patient groups.

RESULTS

Our analysis shows that tumors significantly upregulating genes related to chromosome instability strongly correlate with worse prognosis in breast cancer. Moreover, in multiple tumor types, these tumors upregulate a senescence-bypass transcriptional program and exhibit similar stress phenotypes.

CONCLUSIONS

Using SLEA we are able to find relationships between stress phenotype pathways across multiple cancer types. Moreover we show that SLEA enables the identification of gene sets in correlation with clinical characteristics such as survival, as well as the identification of biological pathways/processes that underlie the pathology of different cancer subgroups.

Associations between HIV and human pathways revealed by protein-protein interactions and correlated gene expression profiles

Background

AIDS is one of the most devastating diseases in human history. Decades of studies have revealed host factors required for HIV infection, indicating that HIV exploits host processes for its own purposes. HIV infection leads to AIDS as well as various comorbidities. The associations between HIV and human pathways and diseases may reveal non-obvious relationships between HIV and non-HIV-defining diseases.

Principal Findings

Human biological pathways were evaluated and statistically compared against the presence of HIV host factor related genes. All of the obtained scores comparing HIV targeted genes and biological pathways were ranked. Different rank results based on overlapping genes, recovered virus-host interactions, co-expressed genes, and common interactions in human protein-protein interaction networks were obtained. Correlations between rankings suggested that these measures yielded diverse rankings. Rank combination of these ranks led to a final ranking of HIV-associated pathways, which revealed that HIV is associated with immune cell-related pathways and several cancer-related pathways. The proposed method is also applicable to the evaluation of associations between other pathogens and human pathways and diseases.

Conclusions

Our results suggest that HIV infection shares common molecular mechanisms with certain signaling pathways and cancers. Interference in apoptosis pathways and the long-term suppression of immune system functions by HIV infection might contribute to tumorigenesis. Relationships between HIV infection and human pathways of disease may aid in the identification of common drug targets for viral infections and other diseases.

Integrating fragmented software applications into holistic solutions: focus on drug discovery

INTRODUCTION

Current advances in software development and global molecular profiling technologies allow the development of holistic software solutions for drug discovery. Such solutions must streamline in silico drug and therapy development by integrating all types of data into one knowledge base and also by enabling continuous analysis workflows uninterrupted by manual restructuring of inputs and outputs from workflow components. They must provide a collaborative environment for data sharing between multiple users and allow importing of all types of experimental data for subsequent analysis.

AREAS COVERED

The reader is provided with a review of disparate software applications currently used in drug development and a discussion of existing organizational challenges for development of holistic software solutions. The reader is also provided with a proposed conceptual framework for integration of software components and some details for its implementation are suggested.

EXPERT OPINION

Holistic solutions can undoubtedly affect the speed, quality and cost of drug development and personalized therapy. However, it must be constantly evolved to rapidly adopt new experimental and statistical methods, incorporate advances in software technologies and allow perpetual optimization of its components. Perpetual improvements in data structure, data quality, statistical algorithms and other mathematical approaches for computer modeling can gradually shift financial and cultural emphasis in the pharmaceutical industry away from traditional experimental approaches and towards computational approaches.

The role of miR-31 and its target gene SATB2 in cancer-associated fibroblasts

It is well established that there is a dynamic relationship between the expanding tumor and the host surrounding tissue. Cancer-associated fibroblasts (CAFs), the most common cellular population found in the tumor microenvironment, supporting tumor growth and dissemination. Here, we set out to determine the factors that may be involved in dramatic alteration of gene expression pattern in CAFs, focusing on microRNA and transcriptional regulators. We established matched pairs of human CAFs isolated from endometrial cancer and normal endometrial fibroblasts. MicroRNA and mRNA analyses identified differential expression of 11 microRNAs, with miR-31 being the most downregulated microRNA in CAFs (p = 0.007). We examined several putative miR-31 target genes identified by microarray analysis and demonstrated that miR-31 directly targets the homeobox gene SATB2, which is responsible for chromatin remodeling and regulation of gene expression, and was significantly elevated in CAFs. The functional relevance of miR-31 and SATB2 were tested in in vitro models of endometrial cancer. Overexpression of miR-31 significantly impaired the ability of CAFs to stimulate tumor cell migration and invasion, without affecting tumor cell proliferation. Genetic manipulation of SATB2 levels in normal fibroblasts or CAFs showed that, reciprocally to miR-31, SATB2 increased tumor cell migration and invasion, while knockdown of endogenous SATB2 in CAFs reversed this phenotype. Introduction of SATB2 into normal fibroblasts stimulated expression of a number of genes involved in cell invasion, migration and scattering. These findings provide new insights into tumor-stroma interaction and document that miR-31 and its target gene SATB2, are involved in regulation of tumor cell motility.

Functional analysis: evaluation of response intensities--tailoring ANOVA for lists of expression subsets

Background

Microarray data is frequently used to characterize the expression profile of a whole genome and to compare the characteristics of that genome under several conditions. Geneset analysis methods have been described previously to analyze the expression values of several genes related by known biological criteria (metabolic pathway, pathology signature, co-regulation by a common factor, etc.) at the same time and the cost of these methods allows for the use of more values to help discover the underlying biological mechanisms.

Results

As several methods assume different null hypotheses, we propose to reformulate the main question that biologists seek to answer. To determine which genesets are associated with expression values that differ between two experiments, we focused on three ad hoc criteria: expression levels, the direction of individual gene expression changes (up or down regulation), and correlations between genes. We introduce the FAERI methodology, tailored from a two-way ANOVA to examine these criteria. The significance of the results was evaluated according to the self-contained null hypothesis, using label sampling or by inferring the null distribution from normally distributed random data. Evaluations performed on simulated data revealed that FAERI outperforms currently available methods for each type of set tested. We then applied the FAERI method to analyze three real-world datasets on hypoxia response. FAERI was able to detect more genesets than other methodologies, and the genesets selected were coherent with current knowledge of cellular response to hypoxia. Moreover, the genesets selected by FAERI were confirmed when the analysis was repeated on two additional related datasets.

Conclusions

The expression values of genesets are associated with several biological effects. The underlying mathematical structure of the genesets allows for analysis of data from several genes at the same time. Focusing on expression levels, the direction of the expression changes, and correlations, we showed that two-step data reduction allowed us to significantly improve the performance of geneset analysis using a modified two-way ANOVA procedure, and to detect genesets that current methods fail to detect.

Evidence for gender-specific transcriptional profiles of nigral dopamine neurons in Parkinson disease

Background

Epidemiological data suggest that the male gender is one of the risks factors for the development of Parkinson Disease (PD). Also, differences in the clinical manifestation and the course of PD have been observed between males and females. However, little is known about the molecular aspects underlying gender-specificity in PD. To address this issue, we determined the gene expression profiles of male and female dopamine (DA) neurons in sporadic PD.

Methodology/Principal Findings

We analyzed Affymetrix-based microarrays on laser microdissected DA neurons from postmortem brains of sporadic PD patients and age-matched controls across genders. Pathway enrichment demonstrated that major cellular pathways involved in PD pathogenesis showed different patterns of deregulation between males and females with more prominent downregulation of genes related to oxidative phosporylation, apoptosis, synaptic transmission and transmission of nerve impulse in the male population. In addition, we found upregulation of gene products for metabolic processes and mitochondrial energy consumption in the age-matched male control neurons. On the single cell level, selected data validation using quantitative Real-Time (qRT)-PCR was consistent with microarray raw data and supported some of the observations from data analysis.

Conclusions/Significance

On the molecular level, our results provide evidence that the expression profiles of aged normal and PD midbrain DA neurons are gender-specific. The observed differences in the expression profiles suggest a disease bias of the male gender, which could be in concordance with clinical observations that the male gender represents a risk factor for sporadic PD. Validation of gene expression by qRT-PCR supported the microarray results, but also pointed to several caveats involved in data interpretation.

Pandora, a pathway and network discovery approach based on common biological evidence

Motivation: Many biological phenomena involve extensive interactions between many of the biological pathways present in cells. However, extraction of all the inherent biological pathways remains a major challenge in systems biology. With the advent of high-throughput functional genomic techniques, it is now possible to infer biological pathways and pathway organization in a systematic way by integrating disparate biological information.

Results: Here, we propose a novel integrated approach that uses network topology to predict biological pathways. We integrated four types of biological evidence (protein–protein interaction, genetic interaction, domain–domain interaction and semantic similarity of Gene Ontology terms) to generate a functionally associated network. This network was then used to develop a new pathway finding algorithm to predict biological pathways in yeast. Our approach discovered 195 biological pathways and 31 functionally redundant pathway pairs in yeast. By comparing our identified pathways to three public pathway databases (KEGG, BioCyc and Reactome), we observed that our approach achieves a maximum positive predictive value of 12.8% and improves on other predictive approaches. This study allows us to reconstruct biological pathways and delineates cellular machinery in a systematic view.

Availability: The method has been implemented in Perl and is available for downloading from http://www.oicr.on.ca/research/ouellette/pandora. It is distributed under the terms of GPL (http://opensource.org/licenses/gpl-2.0.php)

Contact:francis@oicr.on.ca

Supplementary information:Supplementary data are available at Bioinformatics online.

Seeking unique and common biological themes in multiple gene lists or datasets: pathway pattern extraction pipeline for pathway-level comparative analysis

BACKGROUND

One of the challenges in the analysis of microarray data is to integrate and compare the selected (e.g., differential) gene lists from multiple experiments for common or unique underlying biological themes. A common way to approach this problem is to extract common genes from these gene lists and then subject these genes to enrichment analysis to reveal the underlying biology. However, the capacity of this approach is largely restricted by the limited number of common genes shared by datasets from multiple experiments, which could be caused by the complexity of the biological system itself.

RESULTS

We now introduce a new Pathway Pattern Extraction Pipeline (PPEP), which extends the existing WPS application by providing a new pathway-level comparative analysis scheme. To facilitate comparing and correlating results from different studies and sources, PPEP contains new interfaces that allow evaluation of the pathway-level enrichment patterns across multiple gene lists. As an exploratory tool, this analysis pipeline may help reveal the underlying biological themes at both the pathway and gene levels. The analysis scheme provided by PPEP begins with multiple gene lists, which may be derived from different studies in terms of the biological contexts, applied technologies, or methodologies. These lists are then subjected to pathway-level comparative analysis for extraction of pathway-level patterns. This analysis pipeline helps to explore the commonality or uniqueness of these lists at the level of pathways or biological processes from different but relevant biological systems using a combination of statistical enrichment measurements, pathway-level pattern extraction, and graphical display of the relationships of genes and their associated pathways as Gene-Term Association Networks (GTANs) within the WPS platform. As a proof of concept, we have used the new method to analyze many datasets from our collaborators as well as some public microarray datasets.

CONCLUSION

This tool provides a new pathway-level analysis scheme for integrative and comparative analysis of data derived from different but relevant systems. The tool is freely available as a Pathway Pattern Extraction Pipeline implemented in our existing software package WPS, which can be obtained at http://www.abcc.ncifcrf.gov/wps/wps_index.php.