Using the gene ontology for microarray data mining: a comparison of methods and application to age effects in human prefrontal cortex

Down-weighting overlapping genes improves gene set analysis

Background

The identification of gene sets that are significantly impacted in a given condition based on microarray data is a crucial step in current life science research. Most gene set analysis methods treat genes equally, regardless how specific they are to a given gene set.

Results

In this work we propose a new gene set analysis method that computes a gene set score as the mean of absolute values of weighted moderated gene t-scores. The gene weights are designed to emphasize the genes appearing in few gene sets, versus genes that appear in many gene sets. We demonstrate the usefulness of the method when analyzing gene sets that correspond to the KEGG pathways, and hence we called our method Pathway Analysis with Down-weighting of Overlapping Genes (PADOG). Unlike most gene set analysis methods which are validated through the analysis of 2-3 data sets followed by a human interpretation of the results, the validation employed here uses 24 different data sets and a completely objective assessment scheme that makes minimal assumptions and eliminates the need for possibly biased human assessments of the analysis results.

Conclusions

PADOG significantly improves gene set ranking and boosts sensitivity of analysis using information already available in the gene expression profiles and the collection of gene sets to be analyzed. The advantages of PADOG over other existing approaches are shown to be stable to changes in the database of gene sets to be analyzed. PADOG was implemented as an R package available at: http://bioinformaticsprb.med.wayne.edu/PADOG/or http://www.bioconductor.org.

Cancer core modules identification through genomic and transcriptomic changes correlation detection at network level

BACKGROUND

Identification of driver mutations among numerous genomic alternations remains a critical challenge to the elucidation of the underlying mechanisms of cancer. Because driver mutations by definition are associated with a greater number of cancer phenotypes compared to other mutations, we hypothesized that driver mutations could more easily be identified once the genotype-phenotype correlations are detected across tumor samples.

RESULTS

In this study, we describe a novel network analysis to identify the driver mutation through integrating both cancer genomes and transcriptomes. Our method successfully identified a significant genotype-phenotype change correlation in all six solid tumor types and revealed core modules that contain both significantly enriched somatic mutations and aberrant expression changes specific to tumor development. Moreover, we found that the majority of these core modules contained well known cancer driver mutations, and that their mutated genes tended to occur at hub genes with central regulatory roles. In these mutated genes, the majority were cancer-type specific and exhibited a closer relationship within the same cancer type rather than across cancer types. The remaining mutated genes that exist in multiple cancer types led to two cancer type clusters, one cluster consisted of three neural derived or related cancer types, and the other cluster consisted of two adenoma cancer types.

CONCLUSIONS

Our approach can successfully identify the candidate drivers from the core modules. Comprehensive network analysis on the core modules potentially provides critical insights into convergent cancer development in different organs.

Logical Gene Ontology Annotations (GOAL): exploring gene ontology annotations with OWL

MOTIVATION

Ontologies such as the Gene Ontology (GO) and their use in annotations make cross species comparisons of genes possible, along with a wide range of other analytical activities. The bio-ontologies community, in particular the Open Biomedical Ontologies (OBO) community, have provided many other ontologies and an increasingly large volume of annotations of gene products that can be exploited in query and analysis. As many annotations with different ontologies centre upon gene products, there is a possibility to explore gene products through multiple ontological perspectives at the same time. Questions could be asked that link a gene product's function, process, cellular location, phenotype and disease. Current tools, such as AmiGO, allow exploration of genes based on their GO annotations, but not through multiple ontological perspectives. In addition, the semantics of these ontology's representations should be able to, through automated reasoning, afford richer query opportunities of the gene product annotations than is currently possible.

RESULTS

To do this multi-perspective, richer querying of gene product annotations, we have created the Logical Gene Ontology, or GOAL ontology, in OWL that combines the Gene Ontology, Human Disease Ontology and the Mammalian Phenotype Ontology, together with classes that represent the annotations with these ontologies for mouse gene products. Each mouse gene product is represented as a class, with the appropriate relationships to the GO aspects, phenotype and disease with which it has been annotated. We then use defined classes to query these protein classes through automated reasoning, and to build a complex hierarchy of gene products. We have presented this through a Web interface that allows arbitrary queries to be constructed and the results displayed.

CONCLUSION

This standard use of OWL affords a rich interaction with Gene Ontology, Human Disease Ontology and Mammalian Phenotype Ontology annotations for the mouse, to give a fine partitioning of the gene products in the GOAL ontology. OWL in combination with automated reasoning can be effectively used to query across ontologies to ask biologically rich questions. We have demonstrated that automated reasoning can be used to deliver practical on-line querying support for the ontology annotations available for the mouse.

AVAILABILITY

The GOAL Web page is to be found at http://owl.cs.manchester.ac.uk/goal.

Rigorous assessment of gene set enrichment tests

MOTIVATION

Several statistical tests are available to detect the enrichment of differential expression in gene sets. Such tests were originally proposed for analyzing gene sets associated with biological processes. The objective evaluation of tests on real measurements has not been possible as it is difficult to decide a priori, which processes will be affected in given experiments.

RESULTS

We present a first large study to rigorously assess and compare the performance of gene set enrichment tests on real expression measurements. Gene sets are defined based on the targets of given regulators such as transcription factors (TFs) and microRNAs (miRNAs). In contrast to processes, TFs and miRNAs are amenable to direct perturbations, e.g. regulator over-expression or deletion. We assess the ability of 14 different statistical tests to predict the perturbations from expression measurements in Escherichia coli, Saccharomyces cerevisiae and human. We also analyze how performance depends on the quality and comprehensiveness of the regulator targets via a permutation approach. We find that ANOVA and Wilcoxons test consistently perform better than for instance Kolmogorov-Smirnov and hypergeometric tests. For scenarios where the optimal test is not known, we suggest to combine all evaluated tests into an unweighted consensus, which also performs well in our assessment. Our results provide a guide for the selection of existing tests as well as a basis for the development and assessment of novel tests.

Transcriptome Analysis of Cold Syndrome Using Microarray

Microarrays are widely used to study changes in gene expression in diseases. In this paper, we use this technology to discover gene expression patterns in the cold syndrome in Chinese medicine. We identify differentially expressed genes and extracted gene modules that are enriched with differentially expressed genes in the cold syndrome by analyzing cDNA samples, which are purified from blood taken from a pedigree. Our results suggest that the cold syndrome might be caused by the physiological imbalance and/or the disorder of metabolite processes. The study confirms the hypotheses about molecular pathways responsible to human metabolic-related diseases.

Gene expression profile of adult human olfactory bulb and embryonic neural stem cell suggests distinct signaling pathways and epigenetic control

Global gene expression profiling was performed using RNA from human embryonic neural stem cells (hENSC), and adult human olfactory bulb-derived neural stem cells (OBNSCs), to define a gene expression pattern and signaling pathways that are specific for each cell lineage. We have demonstrated large differences in the gene expression profile of human embryonic NSC, and adult human OBNSCs, but less variability between parallel cultures. Transcripts of genes involved in neural tube development and patterning (ALDH1A2, FOXA2), progenitor marker genes (LMX1a, ALDH1A1, SOX10), proliferation of neural progenitors (WNT1 and WNT3a), neuroplastin (NPTN), POU3F1 (OCT6), neuroligin (NLGN4X), MEIS2, and NPAS1 were up-regulated in both cell populations. By Gene Ontology, 325 out of 3875 investigated gene sets were scientifically different. 41 out of the 307 investigated Cellular Component (CC) categories, 45 out of the 620 investigated Molecular Function (MF) categories, and 239 out of the 2948 investigated Biological Process (BP) categories were significant. KEGG Pathway Class Comparison had revealed that 75 out of 171 investigated gene sets passed the 0.005 significance threshold. Levels of gene expression were explored in three signaling pathways, Notch, Wnt, and mTOR that are known to be involved in NS cell fates determination. The transcriptional signature also deciphers the role of genes involved in epigenetic modifications. SWI/SNF DNA chromatin remodeling complex family, including SMARCC1 and SMARCE1, were found specifically up-regulated in our OBNSC but not in hENSC. Differences in gene expression profile of transcripts controlling epigenetic modifications, and signaling pathways might indicate differences in the therapeutic potential of our examined two cell populations in relation to in cell survival, proliferation, migration, and differentiation following engraftments in different CNS insults.

High accordance in prognosis prediction of colorectal cancer across independent datasets by multi-gene module expression profiles

A considerable portion of patients with colorectal cancer have a high risk of disease recurrence after surgery. These patients can be identified by analyzing the expression profiles of signature genes in tumors. But there is no consensus on which genes should be used and the performance of specific set of signature genes varies greatly with different datasets, impeding their implementation in the routine clinical application. Instead of using individual genes, here we identified functional multi-gene modules with significant expression changes between recurrent and recurrence-free tumors, used them as the signatures for predicting colorectal cancer recurrence in multiple datasets that were collected independently and profiled on different microarray platforms. The multi-gene modules we identified have a significant enrichment of known genes and biological processes relevant to cancer development, including genes from the chemokine pathway. Most strikingly, they recruited a significant enrichment of somatic mutations found in colorectal cancer. These results confirmed the functional relevance of these modules for colorectal cancer development. Further, these functional modules from different datasets overlapped significantly. Finally, we demonstrated that, leveraging above information of these modules, our module based classifier avoided arbitrary fitting the classifier function and screening the signatures using the training data, and achieved more consistency in prognosis prediction across three independent datasets, which holds even using very small training sets of tumors.

Ten years of pathway analysis: current approaches and outstanding challenges

Pathway analysis has become the first choice for gaining insight into the underlying biology of differentially expressed genes and proteins, as it reduces complexity and has increased explanatory power. We discuss the evolution of knowledge base–driven pathway analysis over its first decade, distinctly divided into three generations. We also discuss the limitations that are specific to each generation, and how they are addressed by successive generations of methods. We identify a number of annotation challenges that must be addressed to enable development of the next generation of pathway analysis methods. Furthermore, we identify a number of methodological challenges that the next generation of methods must tackle to take advantage of the technological advances in genomics and proteomics in order to improve specificity, sensitivity, and relevance of pathway analysis.

Poly(ADP-ribose) polymerase inhibition enhances p53-dependent and -independent DNA damage responses induced by DNA damaging agent

Targeting DNA repair with poly(ADP-ribose) polymerase (PARP) inhibitors has shown a broad range of anti-tumor activity in patients with advanced malignancies with and without BRCA deficiency. It remains unclear what role p53 plays in response to PARP inhibition in BRCA-proficient cancer cells treated with DNA damaging agents. Using gene expression microarray analysis, we find that DNA damage response (DDR) pathways elicited by veliparib (ABT-888), a PARP inhibitor, plus topotecan comprise the G1/S checkpoint, ATM, and p53 signaling pathways in p53-wildtype cancer cell lines and BRCA1, BRCA2 and ATR pathway in p53-mutant lines. In contrast, topotecan alone induces the G1/S checkpoint pathway in p53-wildtype lines and not in p53-mutant cells. These responses are coupled with G2/G1 checkpoint effectors p21(CDKN1A) upregulation, and Chk1 and Chk2 activation. The drug combination enhances G2 cell cycle arrest, apoptosis and a marked increase in cell death relative to topotecan alone in p53-wildtype and p53-mutant or -null cells. We also show that the checkpoint kinase inhibitor UCN-01 abolishes the G2 arrest induced by the veliparib and topotecan combination and further increases cell death in both p53-wildtype and -mutant cells. Collectively, PARP inhibition by veliparib enhances DDR and cell death in BRCA-proficient cancer cells in a p53-dependent and -independent fashion. Abrogating the cell-cycle arrest induced by PARP inhibition plus chemotherapeutics may be a strategy in the treatment of BRCA-proficient cancer.

Finding consistent disease subnetworks across microarray datasets

Background

While contemporary methods of microarray analysis are excellent tools for studying individual microarray datasets, they have a tendency to produce different results from different datasets of the same disease. We aim to solve this reproducibility problem by introducing a technique (SNet). SNet provides both quantitative and descriptive analysis of microarray datasets by identifying specific connected portions of pathways that are significant. We term such portions within pathways as “subnetworks”.

Results

We tested SNet on independent datasets of several diseases, including childhood ALL, DMD and lung cancer. For each of these diseases, we obtained two independent microarray datasets produced by distinct labs on distinct platforms. In each case, our technique consistently produced almost the same list of significant nontrivial subnetworks from two independent sets of microarray data. The gene-level agreement of these significant subnetworks was between 51.18% to 93.01%. In contrast, when the same pairs of microarray datasets were analysed using GSEA, t-test and SAM, this percentage fell between 2.38% to 28.90% for GSEA, 49.60% tp 73.01% for t-test, and 49.96% to 81.25% for SAM. Furthermore, the genes selected using these existing methods did not form subnetworks of substantial size. Thus it is more probable that the subnetworks selected by our technique can provide the researcher with more descriptive information on the portions of the pathway actually affected by the disease.

Conclusions

These results clearly demonstrate that our technique generates significant subnetworks and genes that are more consistent and reproducible across datasets compared to the other popular methods available (GSEA, t-test and SAM). The large size of subnetworks which we generate indicates that they are generally more biologically significant (less likely to be spurious). In addition, we have chosen two sample subnetworks and validated them with references from biological literature. This shows that our algorithm is capable of generating descriptive biologically conclusions.

A phase II neoadjuvant trial of anastrozole, fulvestrant, and gefitinib in patients with newly diagnosed estrogen receptor positive breast cancer

Endocrine therapy in patients with breast cancer can be limited by the problem of resistance. Preclinical studies suggest that complete blockade of the estrogen receptor (ER) combined with inhibition of the epidermal growth factor receptor can overcome endocrine resistance. We tested this hypothesis in a phase II neoadjuvant trial of anastrozole and fulvestrant combined with gefitinib in postmenopausal women with newly diagnosed ER-positive breast cancer. After a baseline tumor core biopsy, patients were randomized to receive anastrozole and fulvestrant or anastrozole, fulvestrant, and gefitinib (AFG) for 3 weeks. After a second biopsy at 3 weeks, all patients received AFG for 4 months and surgery was done if the tumor was operable. The primary endpoint was best clinical response by RECIST criteria and secondary endpoints were toxicity and change in biomarkers. The study closed after 15 patients were enrolled because of slow accrual. Median patient age was 67 years and median clinical tumor size was 7 cm. Four patients had metastatic disease present. Three patients withdrew before response was assessed. In the remaining 12 patients, there were two complete clinical responses (17%), three partial responses (25%), five had stable disease (41%), and two (17%) had progressive disease. Most common adverse events were rash in four patients, diarrhea in four, joint symptoms in three, and abnormal liver function tests in three. There were no grade 4 toxicities and all toxicities were reversible. At 3 weeks, cell proliferation as measured by Ki-67 was significantly reduced in the AFG group (P value = 0.01), with a parallel reduction in the expression of the Cyclin D1 (P value = 0.02). RNA microarray data showed a corresponding decrease in the expression of cell cycle genes. These results suggest that AFG was an effective neoadjuvant therapy and consistently reduced proliferation in ER-positive tumors.

Impact of aging vs. estrogen loss on cardiac gene expression: estrogen replacement and inflammation

Despite an abundance of evidence to the contrary from animal studies, large clinical trials on humans have shown that estrogen administered to postmenopausal women increases the risk of cardiovascular disease. However, timing may be everything, as estrogen is often administered immediately after ovariectomy (Ovx) in animal studies, while estrogen administration in human studies occurred many years postmenopause. This study investigates the discrepancy by administering 17β-estradiol (E2) in a slow-release capsule to Norway Brown rats both immediately following Ovx and 9 wk post-Ovx (Late), and studying differences in gene expression between these two groups compared with age-matched Ovx and sham-operated animals. Two different types of microarray were used to analyze the left ventricles from these groups: an Affymetrix array (n = 3/group) and an inflammatory cytokines and receptors PCR array (n = 4/group). Key genes were analyzed by Western blotting. Ovx without replacement led to an increase in caspase 3, caspase 9, calpain 2, matrix metalloproteinase (MMP)9, and TNF-α. Caspase 6, STAT3, and CD11b increased in the Late group, while tissue inhibitor of metalloproteinase 2, MMP14, and collagen I α1 were decreased. MADD and fibronectin were increased in both Ovx and Late. TNF-α and inducible nitric oxide synthase (iNOS) protein levels increased with Late replacement. Many of these changes were prevented by early E2 replacement. These findings suggest that increased expression of inflammatory genes, such as TNF-α and iNOS, may be involved in some of the deleterious effects of delayed E2 administration seen in human studies.

TAFFEL: Independent Enrichment Analysis of gene sets

BACKGROUND

A major challenge in genomic research is identifying significant biological processes and generating new hypotheses from large gene sets. Gene sets often consist of multiple separate biological pathways, controlled by distinct regulatory mechanisms. Many of these pathways and the associated regulatory mechanisms might be obscured by a large number of other significant processes and thus not identified as significant by standard gene set enrichment analysis tools.

RESULTS

We present a novel method called Independent Enrichment Analysis (IEA) and software TAFFEL that eases the task by clustering genes to subgroups using Gene Ontology categories and transcription regulators. IEA indicates transcriptional regulators putatively controlling biological functions in studied condition.

CONCLUSIONS

We demonstrate that the developed method and TAFFEL tool give new insight to the analysis of differentially expressed genes and can generate novel hypotheses. Our comparison to other popular methods showed that the IEA method implemented in TAFFEL can find important biological phenomena, which are not reported by other methods.

Comparison of global tests for functional gene sets in two-group designs and selection of potentially effect-causing genes

MOTIVATION

An important object in the analysis of high-throughput genomic data is to find an association between the expression profile of functional gene sets and the different levels of a group response. Instead of multiple testing procedures which focus on single genes, global tests are usually used to detect a group effect in an entire gene set. In a simulation study, we compare the power and computation times of four different approaches for global testing. The applicability of one of these methods to gene expression data is demonstrated for the first time. In addition, we propose an algorithm for the detection of those genes which might be responsible for a group effect.

RESULTS

We could detect that the power of three of the approaches is comparable in many settings but considerable differences were detected in the computation times. Our proposed gene selection algorithm was able to detect potentially effect-causing genes in artificial sets with high power when many genes were altered with a small effect, while classical multiple testing was more powerful when few genes were altered with a large effect.

AVAILABILITY

An R-package called 'RepeatedHighDim' which implements our new global test procedures is made available from http://cran.r-project.org/.

Urine metabolomic analysis identifies potential biomarkers and pathogenic pathways in kidney cancer

Kidney cancer is the seventh most common cancer in the Western world, its incidence is increasing, and it is frequently metastatic at presentation, at which stage patient survival statistics are grim. In addition, there are no useful biofluid markers for this disease, such that diagnosis is dependent on imaging techniques that are not generally used for screening. In the present study, we use metabolomics techniques to identify metabolites in kidney cancer patients' urine, which appear at different levels (when normalized to account for urine volume and concentration) from the same metabolites in nonkidney cancer patients. We found that quinolinate, 4-hydroxybenzoate, and gentisate are differentially expressed at a false discovery rate of 0.26, and these metabolites are involved in common pathways of specific amino acid and energetic metabolism, consistent with high tumor protein breakdown and utilization, and the Warburg effect. When added to four different (three kidney cancer-derived and one "normal") cell lines, several of the significantly altered metabolites, quinolinate, α-ketoglutarate, and gentisate, showed increased or unchanged cell proliferation that was cell line-dependent. Further evaluation of the global metabolomics analysis, as well as confirmation of the specific potential biomarkers using a larger sample size, will lead to new avenues of kidney cancer diagnosis and therapy.

Investigating the effect of paralogs on microarray gene-set analysis

Background

In order to interpret the results obtained from a microarray experiment, researchers often shift focus from analysis of individual differentially expressed genes to analyses of sets of genes. These gene-set analysis (GSA) methods use previously accumulated biological knowledge to group genes into sets and then aim to rank these gene sets in a way that reflects their relative importance in the experimental situation in question. We suspect that the presence of paralogs affects the ability of GSA methods to accurately identify the most important sets of genes for subsequent research.

Results

We show that paralogs, which typically have high sequence identity and similar molecular functions, also exhibit high correlation in their expression patterns. We investigate this correlation as a potential confounding factor common to current GSA methods using Indygene http://www.cbio.uct.ac.za/indygene, a web tool that reduces a supplied list of genes so that it includes no pairwise paralogy relationships above a specified sequence similarity threshold. We use the tool to reanalyse previously published microarray datasets and determine the potential utility of accounting for the presence of paralogs.

Conclusions

The Indygene tool efficiently removes paralogy relationships from a given dataset and we found that such a reduction, performed prior to GSA, has the ability to generate significantly different results that often represent novel and plausible biological hypotheses. This was demonstrated for three different GSA approaches when applied to the reanalysis of previously published microarray datasets and suggests that the redundancy and non-independence of paralogs is an important consideration when dealing with GSA methodologies.

Literature-aided interpretation of gene expression data with the weighted global test

Most methods for the interpretation of gene expression profiling experiments rely on the categorization of genes, as provided by the Gene Ontology (GO) and pathway databases. Due to the manual curation process, such databases are never up-to-date and tend to be limited in focus and coverage. Automated literature mining tools provide an attractive, alternative approach. We review how they can be employed for the interpretation of gene expression profiling experiments. We illustrate that their comprehensive scope aids the interpretation of data from domains poorly covered by GO or alternative databases, and allows for the linking of gene expression with diseases, drugs, tissues and other types of concepts. A framework for proper statistical evaluation of the associations between gene expression values and literature concepts was lacking and is now implemented in a weighted extension of global test. The weights are the literature association scores and reflect the importance of a gene for the concept of interest. In a direct comparison with classical GO-based gene sets, we show that use of literature-based associations results in the identification of much more specific GO categories. We demonstrate the possibilities for linking of gene expression data to patient survival in breast cancer and the action and metabolism of drugs. Coupling with online literature mining tools ensures transparency and allows further study of the identified associations. Literature mining tools are therefore powerful additions to the toolbox for the interpretation of high-throughput genomics data.

HapX positively and negatively regulates the transcriptional response to iron deprivation in Cryptococcus neoformans

The fungal pathogen Cryptococcus neoformans is a major cause of illness in immunocompromised individuals such as AIDS patients. The ability of the fungus to acquire nutrients during proliferation in host tissue and the ability to elaborate a polysaccharide capsule are critical determinants of disease outcome. We previously showed that the GATA factor, Cir1, is a major regulator both of the iron uptake functions needed for growth in host tissue and the key virulence factors such as capsule, melanin and growth at 37°C. We are interested in further defining the mechanisms of iron acquisition from inorganic and host-derived iron sources with the goal of understanding the nutritional adaptation of C. neoformans to the host environment. In this study, we investigated the roles of the HAP3 and HAPX genes in iron utilization and virulence. As in other fungi, the C. neoformans Hap proteins negatively influence the expression of genes encoding respiratory and TCA cycle functions under low-iron conditions. However, we also found that HapX plays both positive and negative roles in the regulation of gene expression, including a positive regulatory role in siderophore transporter expression. In addition, HapX also positively regulated the expression of the CIR1 transcript. This situation is in contrast to the negative regulation by HapX of genes encoding GATA iron regulatory factors in Aspergillus nidulans and Schizosaccharomyces pombe. Although both hapX and hap3 mutants were defective in heme utilization in culture, only HapX made a contribution to virulence, and loss of HapX in a strain lacking the high-affinity iron uptake system did not cause further attenuation of disease. Therefore, HapX appears to have a minimal role during infection of mammalian hosts and instead may be an important regulator of environmental iron uptake functions. Overall, these results indicated that C. neoformans employs multiple strategies for iron acquisition during infection.

Cryptococcus neoformans causes life-threatening central nervous system infections in immunocompromised people such as AIDS patients. The competition for iron between pathogens such as C. neoformans and mammalian hosts is a key aspect of disease outcome. We previously identified and characterized the major iron regulatory protein Cir1 in C. neoformans, as well as proteins for the transport of iron-binding molecules (siderophores) and for high-affinity iron uptake. In this study, we examined the roles of additional regulatory proteins (Hap proteins) in the response to low-iron conditions and the use of host iron sources such as heme and transferrin. We discovered that the HapX protein has a conserved regulatory function to repress iron-dependent functions during iron deprivation, as well as a positive regulatory role for the expression of putative siderophore transporters. A hapX mutant was defective in the use of heme as an iron source in culture but was only modestly attenuated for virulence in mice. This result suggests that additional mechanisms for iron uptake must be available to support C. neoformans proliferation in the host, and that HapX may play an important role in environmental iron acquisition.

Gene expression in response to ionizing radiation and family history of gastric cancer

Genes and molecular pathways involved in familial clustering of gastric cancer have not yet been identified. The purpose of the present study was to investigate gene expression changes in response to a cellular stress, and its link with a positive family history for this neoplasia. To this aim leukocytes of healthy first-degree relatives of gastric cancer patients and controls were challenged in vitro with ionizing radiation and gene expression evaluated 4 h later on microarrays with 1,800 cancer-related genes. Eight genes, mainly involved in signal transduction and cell cycle regulation, were differentially expressed in healthy relatives of gastric cancer cases. Functional class scoring by Gene Ontology classification highlighted two G-protein related pathways, implicated in the proliferation of neoplastic tissue, which were differentially expressed in healthy subjects with positive family history of gastric cancer. The relative expression of 84 genes related to these pathways was examined using the SYBR green-based quantitative real-time PCR. The results confirmed the indication of an involvement of G-protein coupled receptor pathways in GC familiarity provided by microarray analysis. This study indicates a possible association between familiarity for gastric cancer and altered transcriptional response to ionizing radiation.

A decade of systems biology

Systems biology provides a framework for assembling models of biological systems from systematic measurements. Since the field was first introduced a decade ago, considerable progress has been made in technologies for global cell measurement and in computational analyses of these data to map and model cell function. It has also greatly expanded into the translational sciences, with approaches pioneered in yeast now being applied to elucidate human development and disease. Here, we review the state of the field with a focus on four emerging applications of systems biology that are likely to be of particular importance during the decade to follow: (a) pathway-based biomarkers, (b) global genetic interaction maps, (c) systems approaches to identify disease genes, and (d) stem cell systems biology. We also cover recent advances in software tools that allow biologists to explore system-wide models and to formulate new hypotheses. The applications and methods covered in this review provide a set of prime exemplars useful to cell and developmental biologists wishing to apply systems approaches to areas of interest.

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.

Consistency, comprehensiveness, and compatibility of pathway databases

Background

It is necessary to analyze microarray experiments together with biological information to make better biological inferences. We investigate the adequacy of current biological databases to address this need.

Description

Our results show a low level of consistency, comprehensiveness and compatibility among three popular pathway databases (KEGG, Ingenuity and Wikipathways). The level of consistency for genes in similar pathways across databases ranges from 0% to 88%. The corresponding level of consistency for interacting genes pairs is 0%-61%. These three original sources can be assumed to be reliable in the sense that the interacting gene pairs reported in them are correct because they are curated. However, the lack of concordance between these databases suggests each source has missed out many genes and interacting gene pairs.

Conclusions

Researchers will hence find it challenging to obtain consistent pathway information out of these diverse data sources. It is therefore critical to enable them to access these sources via a consistent, comprehensive and unified pathway API. We accumulated sufficient data to create such an aggregated resource with the convenience of an API to access its information. This unified resource can be accessed at http://www.pathwayapi.com.

Genome-wide gene expression changes in genetically stable synthetic and natural wheat allohexaploids

*The present study aims to understand regulation of gene expression in synthetic and natural wheat (Triticum aestivum) allohexaploids, that combines the AB genome of Triticum turgidum and the D genome of Aegilops tauschii; and which we have recently characterized as genetically stable. *We conducted a comprehensive genome-wide analysis of gene expression that allowed characterization of the effect of variability of the D genome progenitor, the intergenerational stability as well as the comparison with natural wheat allohexaploid. We used the Affymetrix GeneChip Wheat Genome Array, on which 55 049 transcripts are represented. *Additive expression was shown to represent the majority of expression regulation in the synthetic allohexaploids, where expression for more than c. 93% of transcripts was equal to the mid-parent value measured from a mixture of parental RNA. This leaves c. 2000 (c. 7%) transcripts, in which expression was nonadditive. No global gene expression bias or dominance towards any of the progenitor genomes was observed whereas high intergenerational stability and low effect of the D genome progenitor variability were revealed. *Our study suggests that gene expression regulation in wheat allohexaploids is established early upon allohexaploidization and highly conserved over generations, as demonstrated by the high similarity of expression with natural wheat allohexaploids.

Muscle transcriptomic profiles in pigs with divergent phenotypes for fatness traits

BACKGROUND

Selection for increasing intramuscular fat content would definitively improve the palatability and juiciness of pig meat as well as the sensorial and organoleptic properties of cured products. However, evidences obtained in human and model organisms suggest that high levels of intramuscular fat might alter muscle lipid and carbohydrate metabolism. We have analysed this issue by determining the transcriptomic profiles of Duroc pigs with divergent phenotypes for 13 fatness traits. The strong aptitude of Duroc pigs to have high levels of intramuscular fat makes them a valuable model to analyse the mechanisms that regulate muscle lipid metabolism, an issue with evident implications in the elucidation of the genetic basis of human metabolic diseases such as obesity and insulin resistance.

RESULTS

Muscle gene expression profiles of 68 Duroc pigs belonging to two groups (HIGH and LOW) with extreme phenotypes for lipid deposition and composition traits have been analysed. Microarray and quantitative PCR analysis showed that genes related to fatty acid uptake, lipogenesis and triacylglycerol synthesis were upregulated in the muscle tissue of HIGH pigs, which are fatter and have higher amounts of intramuscular fat than their LOW counterparts. Paradoxically, lipolytic genes also showed increased mRNA levels in the HIGH group suggesting the existence of a cycle where triacylglycerols are continuously synthesized and degraded. Several genes related to the insulin-signalling pathway, that is usually impaired in obese humans, were also upregulated. Finally, genes related to antigen-processing and presentation were downregulated in the HIGH group.

CONCLUSION

Our data suggest that selection for increasing intramuscular fat content in pigs would lead to a shift but not a disruption of the metabolic homeostasis of muscle cells. Future studies on the post-translational changes affecting protein activity or expression as well as information about protein location within the cell would be needed to to elucidate the effects of lipid deposition on muscle metabolism in pigs.

Research resource: expression profiling reveals unexpected targets and functions of the human steroid receptor RNA activator (SRA) gene

The human steroid receptor RNA activator (SRA) gene encodes both noncoding RNAs (ncRNAs) and protein-generating isoforms. In reporter assays, SRA ncRNA enhances nuclear receptor and myogenic differentiation 1 (MyoD)-mediated transcription but also participates in specific corepressor complexes, serving as a distinct scaffold. That SRA RNA levels might affect some biological functions, such as proliferation, apoptosis, steroidogenesis, and myogenesis, has been reported. However, the breadth of endogenous target genes that might be regulated by SRA RNAs remains largely unknown. To address this, we depleted SRA RNA in two human cancer cell lines with small interfering RNAs and then assayed for changes in gene expression by microarray analyses. The majority of significantly changed genes were reduced upon SRA knockdown, implicating SRA RNAs as endogenous coactivators. Unexpectedly, only a small subset of direct estrogen receptor-alpha target genes was affected in estradiol-treated MCF-7 cells. Eight bona fide SRA downstream target genes were identified (SLC2A3, SLC2A12, CCL20, TGFB2, DIO2, TMEM65, TBL1X, and TMPRSS2), representing entirely novel SRA targets, except for TMPRSS2. These data suggest unanticipated roles for SRA in glucose uptake, cellular signaling, T(3) hormone generation, and invasion/metastasis. SRA depletion in MDA-MB-231 cells reduced invasiveness and expression of some genes critical for this process. Consistent with the knockdown data, overexpressed SRA ncRNA coactivates certain target promoters and may enhance the activity of some coregulatory proteins. This study is a valuable resource because it represents the first genome-wide analysis of a mammalian RNA coregulator.

Multidimensional gene set analysis of genomic data

Understanding the functional implications of changes in gene expression, mutations, etc., is the aim of most genomic experiments. To achieve this, several functional profiling methods have been proposed. Such methods study the behaviour of different gene modules (e.g. gene ontology terms) in response to one particular variable (e.g. differential gene expression). In spite to the wealth of information provided by functional profiling methods, a common limitation to all of them is their inherent unidimensional nature. In order to overcome this restriction we present a multidimensional logistic model that allows studying the relationship of gene modules with different genome-scale measurements (e.g. differential expression, genotyping association, methylation, copy number alterations, heterozygosity, etc.) simultaneously. Moreover, the relationship of such functional modules with the interactions among the variables can also be studied, which produces novel results impossible to be derived from the conventional unidimensional functional profiling methods. We report sound results of gene sets associations that remained undetected by the conventional one-dimensional gene set analysis in several examples. Our findings demonstrate the potential of the proposed approach for the discovery of new cell functionalities with complex dependences on more than one variable.

Expression signature of E2F1 and its associated genes predict superficial to invasive progression of bladder tumors

PURPOSE

In approximately 20% of patients with superficial bladder tumors, the tumors progress to invasive tumors after treatment. Current methods of predicting the clinical behavior of these tumors prospectively are unreliable. We aim to identify a molecular signature that can reliably identify patients with high-risk superficial tumors that are likely to progress to invasive tumors.

PATIENTS AND METHODS

Gene expression data were collected from tumor specimens from 165 patients with bladder cancer. Various statistical methods, including leave-one-out cross-validation methods, were applied to identify a gene expression signature that could predict the likelihood of progression to invasive tumors and to test the robustness of the expression signature in an independent cohort. The robustness of the gene expression signature was validated in an independent (n = 353) cohort.

RESULTS

Supervised analysis of gene expression data revealed a gene expression signature that is strongly associated with invasive bladder tumors. A molecular classifier based on this gene expression signature correctly predicted the likelihood of progression of superficial tumor to invasive tumor.

CONCLUSION

We present a molecular signature that can predict, at diagnosis, the likelihood of bladder cancer progression and, possibly, lead to improvements in patient therapy.

Analysis of multiple sarcoma expression datasets: implications for classification, oncogenic pathway activation and chemotherapy resistance

Background

Diagnosis of soft tissue sarcomas (STS) is challenging. Many remain unclassified (not-otherwise-specified, NOS) or grouped in controversial categories such as malignant fibrous histiocytoma (MFH), with unclear therapeutic value. We analyzed several independent microarray datasets, to identify a predictor, use it to classify unclassifiable sarcomas, and assess oncogenic pathway activation and chemotherapy response.

Methodology/Principal Findings

We analyzed 5 independent datasets (325 tumor arrays). We developed and validated a predictor, which was used to reclassify MFH and NOS sarcomas. The molecular “match” between MFH and their predicted subtypes was assessed using genome-wide hierarchical clustering and Subclass-Mapping. Findings were validated in 15 paraffin samples profiled on the DASL platform. Bayesian models of oncogenic pathway activation and chemotherapy response were applied to individual STS samples. A 170-gene predictor was developed and independently validated (80-85% accuracy in all datasets). Most MFH and NOS tumors were reclassified as leiomyosarcomas, liposarcomas and fibrosarcomas. “Molecular match” between MFH and their predicted STS subtypes was confirmed both within and across datasets. This classification revealed previously unrecognized tissue differentiation lines (adipocyte, fibroblastic, smooth-muscle) and was reproduced in paraffin specimens. Different sarcoma subtypes demonstrated distinct oncogenic pathway activation patterns, and reclassified MFH tumors shared oncogenic pathway activation patterns with their predicted subtypes. These patterns were associated with predicted resistance to chemotherapeutic agents commonly used in sarcomas.

Conclusions/Significance

STS profiling can aid in diagnosis through a predictor tracking distinct tissue differentiation in unclassified tumors, and in therapeutic management via oncogenic pathway activation and chemotherapy response assessment.

Gene set enrichment analysis made simple

Among the many applications of microarray technology, one of the most popular is the identification of genes that are differentially expressed in two conditions. A common statistical approach is to quantify the interest of each gene with a p-value, adjust these p-values for multiple comparisons, choose an appropriate cut-off, and create a list of candidate genes. This approach has been criticised for ignoring biological knowledge regarding how genes work together. Recently a series of methods, that do incorporate biological knowledge, have been proposed. However, the most popular method, gene set enrichment analysis (GSEA), seems overly complicated. Furthermore, GSEA is based on a statistical test known for its lack of sensitivity. In this article we compare the performance of a simple alternative to GSEA. We find that this simple solution clearly outperforms GSEA. We demonstrate this with eight different microarray datasets.

Hepatocellular carcinoma displays distinct DNA methylation signatures with potential as clinical predictors

BACKGROUND

Hepatocellular carcinoma (HCC) is characterized by late detection and fast progression, and it is believed that epigenetic disruption may be the cause of its molecular and clinicopathological heterogeneity. A better understanding of the global deregulation of methylation states and how they correlate with disease progression will aid in the design of strategies for earlier detection and better therapeutic decisions.

METHODS AND FINDINGS

We characterized the changes in promoter methylation in a series of 30 HCC tumors and their respective surrounding tissue and identified methylation signatures associated with major risk factors and clinical correlates. A wide panel of cancer-related gene promoters was analyzed using Illumina bead array technology, and CpG sites were then selected according to their ability to classify clinicopathological parameters. An independent series of HCC tumors and matched surrounding tissue was used for validation of the signatures. We were able to develop and validate a signature of methylation in HCC. This signature distinguished HCC from surrounding tissue and from other tumor types, and was independent of risk factors. However, aberrant methylation of an independent subset of promoters was associated with tumor progression and etiological risk factors (HBV or HCV infection and alcohol consumption). Interestingly, distinct methylation of an independent panel of gene promoters was strongly correlated with survival after cancer therapy.

CONCLUSION

Our study shows that HCC tumors exhibit specific DNA methylation signatures associated with major risk factors and tumor progression stage, with potential clinical applications in diagnosis and prognosis.

Gene expression profiles of HIV-1-infected glia and brain: toward better understanding of the role of astrocytes in HIV-1-associated neurocognitive disorders

Astrocytes are the major cellular component of the central nervous system (CNS), and they play multiple roles in brain development, normal brain function, and CNS responses to pathogens and injury. The functional versatility of astrocytes is linked to their ability to respond to a wide array of biological stimuli through finely orchestrated changes in cellular gene expression. Dysregulation of gene expression programs, generally by chronic exposure to pathogenic stimuli, may lead to dysfunction of astrocytes and contribute to neuropathogenesis. Here, we review studies that employ functional genomics to characterize the effects of HIV-1 and viral pathogenic proteins on cellular gene expression in astrocytes in vitro. We also present the first microarray analysis of primary mouse astrocytes exposed to HIV-1 in culture. In spite of different experimental conditions and microarray platforms used, comparison of the astrocyte array data sets reveals several common gene-regulatory changes that may underlie responses of these cells to HIV-1 and its proteins. We also compared the transcriptional profiles of astrocytes with those obtained in analyses of brain tissues of patients with HIV-1 dementia and macaques infected with simian immunodeficiency virus (SIV). Notably, many of the gene characteristics of responses to HIV-1 in cultured astrocytes were also altered in HIV-1 or SIV-infected brains. Functional genomics, in conjunction with other approaches, may help clarify the role of astrocytes in HIV-1 neuropathogenesis.

A metabolomics approach using juvenile cystic mice to identify urinary biomarkers and altered pathways in polycystic kidney disease

Autosomal dominant polycystic kidney disease (ADPKD) is the most common inherited kidney disease and affects 1 in 1,000 individuals. Ultrasound is most often used to diagnose ADPKD; such a modality is only useful late in the disease after macroscopic cysts are present. There is accumulating evidence suggesting that there are common cellular and molecular mechanisms responsible for cystogenesis in human and murine PKD regardless of the genes mutated, and, in the case of complex metabolomic analysis, the use of a mouse model has distinct advantages for proof of principle over a human study. Therefore, in this study we utilized a urinary metabolomics-based investigation using gas chromatography-time of flight mass spectrometry to demonstrate that the cystic mouse can be discriminated from its wild-type counterpart by urine analysis alone. At day 26 of life, before there is serological evidence of kidney dysfunction, affected mice are distinguishable by urine metabolomic analysis; this finding persists through 45 days until 64 days, at which time body weight differences confound the results. Using functional score analysis and the KEGG pathway database, we identify several biologically relevant metabolic pathways which are altered very early in this disease, the most highly represented being the purine and galactose metabolism pathways. In addition, we identify several specific candidate biomarkers, including allantoic acid and adenosine, which are augmented in the urine of young cystic mice. These markers and pathway components, once extended to human disease, may prove useful as a noninvasive means of diagnosing cystic kidney diseases and to suggest novel therapeutic approaches. Thus, urine metabolomics has great diagnostic potential for cystic renal disorders and deserves further study.

Generation of tumor-specific T lymphocytes using dendritic cell/tumor fusions and anti-CD3/CD28

Adoptive immunotherapy with tumor-specific T cells represents a promising treatment strategy for patients with malignancy. However, the efficacy of T-cell therapy has been limited by the ability to expand tumor-reactive cells with an activated phenotype that effectively target malignant cells. We have developed an anticancer vaccine in which patient-derived tumor cells are fused with autologous dendritic cells (DCs), such that a wide array of tumor antigens are presented in the context of DC-mediated costimulation. In this study, we demonstrate that DC/tumor fusions induce T cells that react with tumor and are dramatically expanded by subsequent ligation of the CD3/CD28 costimulatory complex. These T cells exhibit a predominantly activated phenotype as manifested by an increase in the percentage of cells expressing CD69 and interferon gamma. In addition, the T cells upregulate granzyme B expression and are highly effective in lysing autologous tumor targets. Targeting of tumor-specific antigen was demonstrated by the expansion of T cells with specificity for the MUC1 tetramer. Stimulation with anti-CD3/CD28 followed by DC/tumor fusions or either agent alone failed to result in a similar expansion of tumor-reactive T cells. Consistent with these findings, spectratyping analysis demonstrates selective expansion of T-cell clones as manifested by considerable skewing of the Vbeta repertoire following sequential stimulation with DC/tumor fusions and anti-CD3/CD28. Gene expression analysis was notable for the upregulation of inflammatory pathways. These findings indicate that stimulation with DC/tumor fusions provides a unique platform for subsequent expansion with anti-CD3/CD28 in adoptive T-cell therapy of cancer.

Chromosomal copy number changes of locally advanced rectal cancers treated with preoperative chemoradiotherapy

Standard treatment of rectal cancer patients comprises preoperative chemoradiotherapy followed by radical surgery. However, clinicians are faced with the problem that response rates vary from one individual to another. Predictive biomarkers would therefore be helpful. To identify genomic imbalances that might assist in stratifying tumors into responsive or nonresponsive categories, we used metaphase comparative genomic hybridization to prospectively analyze pretherapeutic biopsies from 42 patients with locally advanced rectal cancers. These patients were subsequently treated with 5-fluorouracil-based preoperative chemoradiotherapy. Based on downsizing of the T-category, 21 rectal cancers were later classified as responsive, while the other 21 were nonresponsive. Comparing these two groups, we could show that gains of chromosomal regions 7q32 approximately q36 and 7q11 approximately q31, as well as amplifications of 20q11 approximately q13, were significantly associated with responsiveness to preoperative chemoradiotherapy (P<0.05). However, the probability of detecting these copy number changes by chance is high (P=0.21). Our primary results suggest that pretherapeutic evaluation of chromosomal copy number changes may be of value for response prediction of rectal cancers to preoperative chemoradiotherapy. This will require validation in a larger cohort of patients.

Impact of acute psychosocial stress on peripheral blood gene expression pathways in healthy men

We investigated peripheral blood mononuclear cell gene expression responses to acute psychosocial stress to identify molecular pathways relevant to the stress response. Blood samples were obtained from 10 healthy male subjects before, during and after (at 0, 30, and 60 min) a standardized psychosocial laboratory stressor. Ribonucleic acid (RNA) was extracted and gene expression measured by hybridization to a 20,000-gene microarray. Gene Set Expression Comparisons (GSEC) using defined pathways were used for the analysis. Forty-nine pathways were significantly changed from baseline to immediately after the stressor (p < 0.05), implicating cell cycle, cell signaling, adhesion and immune responses. The comparison between stress and recovery (measured 30 min later) identified 36 pathways, several involving stress-responsive signaling cascades and cellular defense mechanisms. These results have relevance for understanding molecular mechanisms of the physiological stress response, and might be used to further study adverse health outcomes of psychosocial stress.

Text-based over-representation analysis of microarray gene lists with annotation bias

A major challenge in microarray data analysis is the functional interpretation of gene lists. A common approach to address this is over-representation analysis (ORA), which uses the hypergeometric test (or its variants) to evaluate whether a particular functionally defined group of genes is represented more than expected by chance within a gene list. Existing applications of ORA have been largely limited to pre-defined terminologies such as GO and KEGG. We report our explorations of whether ORA can be applied to a wider mining of free-text. We found that a hitherto underappreciated feature of experimentally derived gene lists is that the constituents have substantially more annotation associated with them, as they have been researched upon for a longer period of time. This bias, a result of patterns of research activity within the biomedical community, is a major problem for classical hypergeometric test-based ORA approaches, which cannot account for such bias. We have therefore developed three approaches to overcome this bias, and demonstrate their usability in a wide range of published datasets covering different species. A comparison with existing tools that use GO terms suggests that mining PubMed abstracts can reveal additional biological insight that may not be possible by mining pre-defined ontologies alone.

QTL global meta-analysis: are trait determining genes clustered?

Background

A key open question in biology is if genes are physically clustered with respect to their known functions or phenotypic effects. This is of particular interest for Quantitative Trait Loci (QTL) where a QTL region could contain a number of genes that contribute to the trait being measured.

Results

We observed a significant increase in gene density within QTL regions compared to non-QTL regions and/or the entire bovine genome. By grouping QTL from the Bovine QTL Viewer database into 8 categories of non-redundant regions, we have been able to analyze gene density and gene function distribution, based on Gene Ontology (GO) with relation to their location within QTL regions, outside of QTL regions and across the entire bovine genome. We identified a number of GO terms that were significantly over represented within particular QTL categories. Furthermore, select GO terms expected to be associated with the QTL category based on common biological knowledge have also proved to be significantly over represented in QTL regions.

Conclusion

Our analysis provides evidence of over represented GO terms in QTL regions. This increased GO term density indicates possible clustering of gene functions within QTL regions of the bovine genome. Genes with similar functions may be grouped in specific locales and could be contributing to QTL traits. Moreover, we have identified over-represented GO terminology that from a biological standpoint, makes sense with respect to QTL category type.

Text-mining approach to evaluate terms for ontology development

Developing ontologies to account for the complexity of biological systems requires the time intensive collaboration of many participants with expertise in various fields. While each participant may contribute to construct a list of terms for ontology development, no objective methods have been developed to evaluate how relevant each of these terms is to the intended domain. We have developed a computational method based on a hypergeometric enrichment test to evaluate the relevance of such terms to the intended domain. The proposed method uses the PubMed literature database to evaluate whether each potential term for ontology development is overrepresented in the abstracts that discuss the particular domain. This evaluation provides an objective approach to assess terms and prioritize them for ontology development.

Multivariate analysis of variance test for gene set analysis

MOTIVATION

Gene class testing (GCT) or gene set analysis (GSA) is a statistical approach to determine whether some functionally predefined sets of genes express differently under different experimental conditions. Shortcomings of the Fisher's exact test for the overrepresentation analysis are illustrated by an example. Most alternative GSA methods are developed for data collected from two experimental conditions, and most is based on a univariate gene-by-gene test statistic or assume independence among genes in the gene set. A multivariate analysis of variance (MANOVA) approach is proposed for studies with two or more experimental conditions.

RESULTS

When the number of genes in the gene set is greater than the number of samples, the sample covariance matrix is singular and ill-condition. The use of standard multivariate methods can result in biases in the analysis. The proposed MANOVA test uses a shrinkage covariance matrix estimator for the sample covariance matrix. The MANOVA test and six other GSA published methods, principal component analysis, SAM-GS, analysis of covariance, Global, GSEA and MaxMean, are evaluated using simulation. The MANOVA test appears to perform the best in terms of control of type I error and power under the models considered in the simulation. Several publicly available microarray datasets under two and three experimental conditions are analyzed for illustrations of GSA. Most methods, except for GSEA and MaxMean, generally are comparable in terms of power of identification of significant gene sets.

AVAILABILITY

A free R-code to perform MANOVA test is available at http://mail.cmu.edu.tw/~catsai/research.htm.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Decreased TGFbeta signaling and increased COX2 expression in high risk women with increased mammographic breast density

High mammographic density is associated with a increased risk of breast cancer. We hypothesized that specific pathways exist that are associated with increased mammographic density, and may therefore be used to identify potential targets for chemoprevention. Histologically confirmed normal breast tissue was collected from women undergoing breast surgery who had available demographic data and mammograms for review. Women with low versus high mammographic breast density were compared. Differentially expressed genes using Affymetrix HG U133Plus2 chips were identified in dense versus non-dense tissue. Immunohistochemical analysis (IHC) of estrogen receptor, progesterone receptor, Ki67, and COX2 expression was performed. About 66 women were identified, 28 (42%) had high, and 38 (58%) had low mammographic density. About 73 genes had differential expression between normal breast tissue with high and low mammographic density (P < 0.001, fold change > or = 1.5 with a low false discovery rate (<10%). Network and canonical pathway analysis indicated decreased TGFbeta signaling (TGFBR2, SOS, SMAD3, CD44 and TNFRSF11B) in dense breast tissue relative to non-dense breast. By IHC, only COX2 expression in the stroma was statistically significant on multivariate analysis. TGFbeta ligands are currently the only growth factors known to prevent mammary epithelial cell proliferation. TGFbeta signaling has been reported to be inhibited by COX-2, and these molecules are highly differentially expressed in individuals at high risk of developing breast cancer. These results strongly suggest that COX2 inhibition should be investigated for breast cancer prevention despite possible increase in cardiovascular risk.

Altered proliferation and differentiation properties of primary mammary epithelial cells from BRCA1 mutation carriers

Female BRCA1 mutation carriers have a nearly 80% probability of developing breast cancer during their life-time. We hypothesized that the breast epithelium at risk in BRCA1 mutation carriers harbors mammary epithelial cells (MEC) with altered proliferation and differentiation properties. Using a three-dimensional culture technique to grow MECs ex vivo, we found that the ability to form colonies, an indication of clonality, was restricted to the aldehyde dehydrogenase 1-positive fraction in MECs but not in HCC1937 BRCA1-mutant cancer cells. Primary MECs from BRCA1 mutation carriers (n = 9) had a 28% greater ability for clonal growth compared with normal controls (n = 6; P = 0.006), and their colonies were significantly larger. Colonies in controls and BRCA1 mutation carriers stained positive for BRCA1 by immunohistochemistry, and 79% of the examined single colonies from BRCA1 carriers retained heterozygosity for BRCA1 (ROH). Colonies from BRCA1 mutation carriers frequently showed high epidermal growth factor receptor (EGFR) expression (71% EGFR positive versus 44% in controls) and were negative for estrogen receptor (ERalpha; 32% ER negative, 44% mixed, 24% ER positive versus 90% ER positive in controls). Expression of CK14 and p63 were not significantly different. Microarray studies revealed that colonies from BRCA1-mutant PMECs anticipate expression profiles found in BRCA1-related tumors, and that the EGFR pathway is up-regulated. We conclude that BRCA1 haploinsufficiency leads to an increased ability for clonal growth and proliferation in the PMECs of BRCA1 mutation carriers, possibly as a result of EGFR pathway activation. These altered growth and differentiation properties may render BRCA1-mutant PMECs vulnerable to transformation and predispose to the development of ER-negative, EGFR-positive breast cancers.

Survival-related profile, pathways, and transcription factors in ovarian cancer

BACKGROUND

Ovarian cancer has a poor prognosis due to advanced stage at presentation and either intrinsic or acquired resistance to classic cytotoxic drugs such as platinum and taxoids. Recent large clinical trials with different combinations and sequences of classic cytotoxic drugs indicate that further significant improvement in prognosis by this type of drugs is not to be expected. Currently a large number of drugs, targeting dysregulated molecular pathways in cancer cells have been developed and are introduced in the clinic. A major challenge is to identify those patients who will benefit from drugs targeting these specific dysregulated pathways.The aims of our study were (1) to develop a gene expression profile associated with overall survival in advanced stage serous ovarian cancer, (2) to assess the association of pathways and transcription factors with overall survival, and (3) to validate our identified profile and pathways/transcription factors in an independent set of ovarian cancers.

METHODS AND FINDINGS

According to a randomized design, profiling of 157 advanced stage serous ovarian cancers was performed in duplicate using approximately 35,000 70-mer oligonucleotide microarrays. A continuous predictor of overall survival was built taking into account well-known issues in microarray analysis, such as multiple testing and overfitting. A functional class scoring analysis was utilized to assess pathways/transcription factors for their association with overall survival. The prognostic value of genes that constitute our overall survival profile was validated on a fully independent, publicly available dataset of 118 well-defined primary serous ovarian cancers. Furthermore, functional class scoring analysis was also performed on this independent dataset to assess the similarities with results from our own dataset. An 86-gene overall survival profile discriminated between patients with unfavorable and favorable prognosis (median survival, 19 versus 41 mo, respectively; permutation p-value of log-rank statistic = 0.015) and maintained its independent prognostic value in multivariate analysis. Genes that composed the overall survival profile were also able to discriminate between the two risk groups in the independent dataset. In our dataset 17/167 pathways and 13/111 transcription factors were associated with overall survival, of which 16 and 12, respectively, were confirmed in the independent dataset.

CONCLUSIONS

Our study provides new clues to genes, pathways, and transcription factors that contribute to the clinical outcome of serous ovarian cancer and might be exploited in designing new treatment strategies.

Challenges and approaches to statistical design and inference in high-dimensional investigations

Advances in modern technologies have facilitated high-dimensional experiments (HDEs) that generate tremendous amounts of genomic, proteomic, and other "omic" data. HDEs involving whole-genome sequences and polymorphisms, expression levels of genes, protein abundance measurements, and combinations thereof have become a vanguard for new analytic approaches to the analysis of HDE data. Such situations demand creative approaches to the processes of statistical inference, estimation, prediction, classification, and study design. The novel and challenging biological questions asked from HDE data have resulted in many specialized analytic techniques being developed. This chapter discusses some of the unique statistical challenges facing investigators studying high-dimensional biology and describes some approaches being developed by statistical scientists. We have included some focus on the increasing interest in questions involving testing multiple propositions simultaneously, appropriate inferential indicators for the types of questions biologists are interested in, and the need for replication of results across independent studies, investigators, and settings. A key consideration inherent throughout is the challenge in providing methods that a statistician judges to be sound and a biologist finds informative.

A resampling-based meta-analysis for detection of differential gene expression in breast cancer

Background

Accuracy in the diagnosis of breast cancer and classification of cancer subtypes has improved over the years with the development of well-established immunohistopathological criteria. More recently, diagnostic gene-sets at the mRNA expression level have been tested as better predictors of disease state. However, breast cancer is heterogeneous in nature; thus extraction of differentially expressed gene-sets that stably distinguish normal tissue from various pathologies poses challenges. Meta-analysis of high-throughput expression data using a collection of statistical methodologies leads to the identification of robust tumor gene expression signatures.

Methods

A resampling-based meta-analysis strategy, which involves the use of resampling and application of distribution statistics in combination to assess the degree of significance in differential expression between sample classes, was developed. Two independent microarray datasets that contain normal breast, invasive ductal carcinoma (IDC), and invasive lobular carcinoma (ILC) samples were used for the meta-analysis. Expression of the genes, selected from the gene list for classification of normal breast samples and breast tumors encompassing both the ILC and IDC subtypes were tested on 10 independent primary IDC samples and matched non-tumor controls by real-time qRT-PCR. Other existing breast cancer microarray datasets were used in support of the resampling-based meta-analysis.

Results

The two independent microarray studies were found to be comparable, although differing in their experimental methodologies (Pearson correlation coefficient, R = 0.9389 and R = 0.8465 for ductal and lobular samples, respectively). The resampling-based meta-analysis has led to the identification of a highly stable set of genes for classification of normal breast samples and breast tumors encompassing both the ILC and IDC subtypes. The expression results of the selected genes obtained through real-time qRT-PCR supported the meta-analysis results.

Conclusion

The proposed meta-analysis approach has the ability to detect a set of differentially expressed genes with the least amount of within-group variability, thus providing highly stable gene lists for class prediction. Increased statistical power and stringent filtering criteria used in the present study also make identification of novel candidate genes possible and may provide further insight to improve our understanding of breast cancer development.

Tools for interpreting large-scale protein profiling in microbiology

Quantitative proteomic analysis of microbial systems generates large datasets that can be difficult and time-consuming to interpret. Fortunately, many of the data display and gene-clustering tools developed to analyze large transcriptome microarray datasets are also applicable to proteomes. Plots of abundance ratio vs. total signal or spectral counts can highlight regions of random error and putative change. Displaying data in the physical order of the genes in the genome sequence can highlight potential operons. At a basic level of transcriptional organization, identifying operons can give insights into regulatory pathways as well as provide corroborating evidence for proteomic results. Classification and clustering algorithms can group proteins together by their abundance changes under different conditions, helping to identify interesting expression patterns, but often work poorly with noisy data such as typically generated in a large-scale proteomic analysis. Biological interpretation can be aided more directly by overlaying differential protein abundance data onto metabolic pathways, indicating pathways with altered activities. More broadly, ontology tools detect altered levels of protein abundance for different metabolic pathways, molecular functions, and cellular localizations. In practice, pathway analysis and ontology are limited by the level of database curation associated with the organism of interest.

The cerebral microvasculature in schizophrenia: a laser capture microdissection study

Background

Previous studies of brain and peripheral tissues in schizophrenia patients have indicated impaired energy supply to the brain. A number of studies have also demonstrated dysfunction of the microvasculature in schizophrenia patients. Together these findings are consistent with a hypothesis of blood-brain barrier dysfunction in schizophrenia. In this study, we have investigated the cerebral vascular endothelium of schizophrenia patients at the level of transcriptomics.

Methodology/Principal Findings

We used laser capture microdissection to isolate both microvascular endothelial cells and neurons from post mortem brain tissue from schizophrenia patients and healthy controls. RNA was isolated from these cell populations, amplified, and analysed using two independent microarray platforms, Affymetrix HG133plus2.0 GeneChips and CodeLink Whole Human Genome arrays. In the first instance, we used the dataset to compare the neuronal and endothelial data, in order to demonstrate that the predicted differences between cell types could be detected using this methodology. We then compared neuronal and endothelial data separately between schizophrenic subjects and controls. Analysis of the endothelial samples showed differences in gene expression between schizophrenics and controls which were reproducible in a second microarray platform. Functional profiling revealed that these changes were primarily found in genes relating to inflammatory processes.

Conclusions/Significance

This study provides preliminary evidence of molecular alterations of the cerebral microvasculature in schizophrenia patients, suggestive of a hypo-inflammatory state in this tissue type. Further investigation of the blood-brain barrier in schizophrenia is warranted.

Carboplatin-induced gene expression changes in vitro are prognostic of survival in epithelial ovarian cancer

Background

We performed a time-course microarray experiment to define the transcriptional response to carboplatin in vitro, and to correlate this with clinical outcome in epithelial ovarian cancer (EOC). RNA was isolated from carboplatin and control-treated 36M2 ovarian cancer cells at several time points, followed by oligonucleotide microarray hybridization. Carboplatin induced changes in gene expression were assessed at the single gene as well as at the pathway level. Clinical validation was performed in publicly available microarray datasets using disease free and overall survival endpoints.

Results

Time-course and pathway analyses identified 317 genes and 40 pathways (designated time-course and pathway signatures) deregulated following carboplatin exposure. Both types of signatures were validated in two separate platinum-treated ovarian and NSCLC cell lines using published microarray data. Expression of time-course and pathway signature genes distinguished between patients with unfavorable and favorable survival in two independent ovarian cancer datasets. Among the pathways most highly induced by carboplatin in vitro, the NRF2, NF-kB, and cytokine and inflammatory response pathways were also found to be upregulated prior to chemotherapy exposure in poor prognosis tumors.

Conclusion

Dynamic assessment of gene expression following carboplatin exposure in vitro can identify both genes and pathways that are correlated with clinical outcome. The functional relevance of this observation for better understanding the mechanisms of drug resistance in EOC will require further evaluation.

A novel signaling pathway impact analysis

MOTIVATION

Gene expression class comparison studies may identify hundreds or thousands of genes as differentially expressed (DE) between sample groups. Gaining biological insight from the result of such experiments can be approached, for instance, by identifying the signaling pathways impacted by the observed changes. Most of the existing pathway analysis methods focus on either the number of DE genes observed in a given pathway (enrichment analysis methods), or on the correlation between the pathway genes and the class of the samples (functional class scoring methods). Both approaches treat the pathways as simple sets of genes, disregarding the complex gene interactions that these pathways are built to describe.

RESULTS

We describe a novel signaling pathway impact analysis (SPIA) that combines the evidence obtained from the classical enrichment analysis with a novel type of evidence, which measures the actual perturbation on a given pathway under a given condition. A bootstrap procedure is used to assess the significance of the observed total pathway perturbation. Using simulations we show that the evidence derived from perturbations is independent of the pathway enrichment evidence. This allows us to calculate a global pathway significance P-value, which combines the enrichment and perturbation P-values. We illustrate the capabilities of the novel method on four real datasets. The results obtained on these data show that SPIA has better specificity and more sensitivity than several widely used pathway analysis methods.

AVAILABILITY

SPIA was implemented as an R package available at http://vortex.cs.wayne.edu/ontoexpress/