Strategies for comparing gene expression profiles from different microarray platforms: Application to a case–control experiment

Bioinformatic analysis of differentially expressed genes as prognostic markers in pheochromocytoma and paraganglioma tumors

The pathogenesis of pheochromocytoma and paraganglioma (PCPG) catecholamine-producing tumors is exceedingly complicated. Here, we sought to identify important genes affecting the prognosis and survival rate of patients suffering from PCPG. We analyzed 95 samples obtained from two microarray data series, GSE19422 and GSE60459, from the Gene Expression Omnibus (GEO) repository. First, differentially expressed genes (DEGs) were identified by comparing 87 PCPG tumor samples and eight normal adrenal tissue samples using R language. The GEO2R tool and Venn diagram software were applied to the Database for Annotation, Visualization and Integrated Discovery (DAVID) to analyze Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Gene Ontology (GO). We further employed Cytoscape with the Molecular Complex Detection (MCODE) tool to make protein-protein interactions visible for the Search Tool for Retrieval of Interacting Genes (STRING). These procedures resulted in 30 candidate DEGs, which were subjected to Kaplan-Meier analysis and validated by Gene Expression Profiling Interactive Analysis (GEPIA) to determine their influence on overall survival rate. Finally, we identified ALDH3A2 and AKR1B1, two genes in the glycerolipid metabolism pathway, as being particularly enriched in PCPG tumors and correlated with T and B tumor-infiltrating immune cells. Our results suggest that these two DEGs are closely associated with the prognosis of malignant PCPG tumors.

miRNA-target network reveals miR-124as a key miRNA contributing to clear cell renal cell carcinoma aggressive behaviour by targeting CAV1 and FLOT1

Clear cell renal cell carcinoma (ccRCC) is an aggressive tumor with frequent metastatic rate and poor survival. Integrated analyses allow understanding the interplay between different levels of molecular alterations.

We integrated miRNA and gene expression data from 458 ccRCC and 254 normal kidney specimens to construct a miRNA-target interaction network.

We identified the downregulated miR-124-3p, -30a-5p and -200c-3p as the most influential miRNAs in RCC pathogenesis.miR-124-3p and miR-200c-3p expression showed association with patient survival, miR-30a-5p was downregulated in metastases compared to primary tumors. We used an independent set of 87 matched samples for validation. We confirmed the functional impact of these miRNAs by in vitro assays. Restoration of these miRNAs reduced migration, invasion and proliferation. miR-124-3p decreased the S phase of cell cycle, as well. We compared transcriptome profiling before and after miRNA overexpression, and validated CAV1 and FLOT1 as miR-124-3p targets. Patients with higher CAV1 and FLOT1 had lower miR-124-3p expression and shorter overall survival.

We hypothesize that these three miRNAs are fundamental contributing to ccRCC aggressive/metastatic behavior; and miR-124-3p especially has a key role through regulating CAV1 and FLOT1 expression. Restoration of the levels of these miRNAs could be considered as a potential therapeutic strategy for ccRCC.

Integrative bioinformatics analysis reveals new prognostic biomarkers of clear cell renal cell carcinoma

BACKGROUND

The outcome of clear cell renal cell carcinoma (ccRCC) is still unpredictable. Even with new targeted therapies, the average progression-free survival is dismal. Markers for early detection and progression could improve disease outcome.

METHODS

To identify efficient and hitherto unrecognized pathogenic factors of the disease, we performed a uniquely comprehensive pathway analysis and built a gene interaction network based on large publicly available data sets assembled from 28 publications, comprising a 3-prong approach with high-throughput mRNA, microRNA, and protein expression profiles of 593 ccRCC and 389 normal kidney samples. We validated our results on 2 different data sets of 882 ccRCC and 152 normal tissues. Functional analyses were done by proliferation, migration, and invasion assays following siRNA (small interfering RNA) knockdown.

RESULTS

After integration of multilevel data, we identified aryl-hydrocarbon receptor (AHR), grainyhead-like-2 (GRHL2), and KIAA0101 as new pathogenic factors. GRHL2 expression was associated with higher chances for disease relapse and retained prognostic utility after controlling for grade and stage [hazard ratio (HR), 3.47, P = 0.012]. Patients with KIAA0101-positive expression suffered worse disease-free survival (HR, 3.64, P < 0.001), and in multivariate analysis KIAA0101 retained its independent prognostic significance. Survival analysis showed that GRHL2- and KIAA0101-positive patients had significantly lower disease-free survival (P = 0.002 and P < 0.001). We also found that KIAA0101 silencing decreased kidney cancer cell migration and invasion in vitro.

CONCLUSIONS

Using an integrative system biology approach, we identified 3 novel factors as potential biomarkers (AHR, GRHL2 and KIAA0101) involved in ccRCC pathogenesis and not linked to kidney cancer before.

Supporting regenerative medicine by integrative dimensionality reduction

OBJECTIVE

The assessment of the developmental potential of stem cells is a crucial step towards their clinical application in regenerative medicine. It has been demonstrated that genome-wide expression profiles can predict the cellular differentiation stage by means of dimensionality reduction methods. Here we show that these techniques can be further strengthened to support decision making with i) a novel strategy for gene selection; ii) methods for combining the evidence from multiple data sets.

METHODS

We propose to exploit dimensionality reduction methods for the selection of genes specifically activated in different stages of differentiation. To obtain an integrated predictive model, the expression values of the selected genes from multiple data sets are combined. We investigated distinct approaches that either aggregate data sets or use learning ensembles.

RESULTS

We analyzed the performance of the proposed methods on six publicly available data sets. The selection procedure identified a reduced subset of genes whose expression values gave rise to an accurate stage prediction. The assessment of predictive accuracy demonstrated a high quality of predictions for most of the data integration methods presented.

CONCLUSION

The experimental results highlighted the main potentials of proposed approaches. These include the ability to predict the true staging by combining multiple training data sets when this could not be inferred from a single data source, and to focus the analysis on a reduced list of genes of similar predictive performance.

Intra- and inter-individual variance of gene expression in clinical studies

Background

Variance in microarray studies has been widely discussed as a critical topic on the identification of differentially expressed genes; however, few studies have addressed the influence of estimating variance.

Methodology/Principal Findings

To break intra- and inter-individual variance in clinical studies down to three levels–technical, anatomic, and individual–we designed experiments and algorithms to investigate three forms of variances. As a case study, a group of “inter-individual variable genes” were identified to exemplify the influence of underestimated variance on the statistical and biological aspects in identification of differentially expressed genes. Our results showed that inadequate estimation of variance inevitably led to the inclusion of non-statistically significant genes into those listed as significant, thereby interfering with the correct prediction of biological functions. Applying a higher cutoff value of fold changes in the selection of significant genes reduces/eliminates the effects of underestimated variance.

Conclusions/Significance

Our data demonstrated that correct variance evaluation is critical in selecting significant genes. If the degree of variance is underestimated, “noisy” genes are falsely identified as differentially expressed genes. These genes are the noise associated with biological interpretation, reducing the biological significance of the gene set. Our results also indicate that applying a higher number of fold change as the selection criteria reduces/eliminates the differences between distinct estimations of variance.

The impact of network biology in pharmacology and toxicology

With the need to investigate alternative approaches and emerging technologies in order to increase drug efficacy and reduce adverse drug effects, network biology offers a novel way of approaching drug discovery by considering the effect of a molecule and protein's function in a global physiological environment. By studying drug action across multiple scales of complexity, from molecular to cellular and tissue level, network-based computational methods have the potential to improve our understanding of the impact of chemicals in human health. In this review we present the available large-scale databases and tools that allow integration and analysis of such information for understanding the properties of small molecules in the context of cellular networks. With the recent advances in the omics area, global integrative approaches are necessary to cope with the massive amounts of data, and biomedical researchers are urged to implement new types of analyses that can lead to new therapeutic interventions with increased safety and efficacy compared with existing medications.

A PRACTICAL QUESTION BASED ON CROSS-PLATFORM MICROARRAY DATA NORMALIZATION: ARE BOEC MORE LIKE LARGE VESSEL OR MICROVASCULAR ENDOTHELIAL CELLS OR NEITHER OF THEM?

Since the available microarray data of BOEC (human blood outgrowth endothelial cells), large vessel, and microvascular endothelial cells were from two different platforms, a working cross-platform normalization method was needed to make these data comparable. With six HUVEC (human umbilical vein endothelial cells) samples hybridized on two-channel cDNA arrays and six HUVEC samples on Affymetrix arrays, 64 possible combinations of a three-step normalization procedure were investigated to search for the best normalization method, which was selected, based on two criteria measuring the extent to which expression profiles of biological samples of the same cell type arrayed on two platforms were indistinguishable. Next, three discriminative gene lists between the large vessel and the microvascular endothelial cells were achieved by SAM (significant analysis of microarrays), PAM (prediction analysis for microarrays), and a combination of SAM and PAM lists. The final discriminative gene list was selected by SVM (support vector machine). Based on this discriminative gene list, SVM classification analysis with best tuning parameters and 10,000 times of validations showed that BOEC were far from large vessel cells, they either formed their own class, or fell into the microvascular class. Based on all the common genes between the two platforms, SVM analysis further confirmed this conclusion.

High-throughput processing and normalization of one-color microarrays for transcriptional meta-analyses

Background

Microarray experiments are becoming increasingly common in biomedical research, as is their deposition in publicly accessible repositories, such as Gene Expression Omnibus (GEO). As such, there has been a surge in interest to use this microarray data for meta-analytic approaches, whether to increase sample size for a more powerful analysis of a specific disease (e.g. lung cancer) or to re-examine experiments for reasons different than those examined in the initial, publishing study that generated them. For the average biomedical researcher, there are a number of practical barriers to conducting such meta-analyses such as manually aggregating, filtering and formatting the data. Methods to automatically process large repositories of microarray data into a standardized, directly comparable format will enable easier and more reliable access to microarray data to conduct meta-analyses.

Methods

We present a straightforward, simple but robust against potential outliers method for automatic quality control and pre-processing of tens of thousands of single-channel microarray data files. GEO GDS files are quality checked by comparing parametric distributions and quantile normalized to enable direct comparison of expression level for subsequent meta-analyses.

Results

13,000 human 1-color experiments were processed to create a single gene expression matrix that subsets can be extracted from to conduct meta-analyses. Interestingly, we found that when conducting a global meta-analysis of gene-gene co-expression patterns across all 13,000 experiments to predict gene function, normalization had minimal improvement over using the raw data.

Conclusions

Normalization of microarray data appears to be of minimal importance on analyses based on co-expression patterns when the sample size is on the order of thousands microarray datasets. Smaller subsets, however, are more prone to aberrations and artefacts, and effective means of automating normalization procedures not only empowers meta-analytic approaches, but aids in reproducibility by providing a standard way of approaching the problem.

Data availability: matrix containing normalized expression of 20,813 genes across 13,000 experiments is available for download at . Source code for GDS files pre-processing is available from the authors upon request.

Microarray evidences the role of pathologic adipose tissue in insulin resistance and their clinical implications

Clustering of insulin resistance and dysmetabolism with obesity is attributed to pathologic adipose tissue. The morphologic hallmarks of this pathology are adipocye hypertrophy and heightened inflammation. However, it's underlying molecular mechanisms remains unknown. Study of gene function in metabolically active tissues like adipose tissue, skeletal muscle and liver is a promising strategy. Microarray is a powerful technique of assessment of gene function by measuring transcription of large number of genes in an array. This technique has several potential applications in understanding pathologic adipose tissue. They are: (1) transcriptomic differences between various depots of adipose tissue, adipose tissue from obese versus lean individuals, high insulin resistant versus low insulin resistance, brown versus white adipose tissue, (2) transcriptomic profiles of various stages of adipogenesis, (3) effect of diet, cytokines, adipokines, hormones, environmental toxins and drugs on transcriptomic profiles, (4) influence of adipokines on transcriptomic profiles in skeletal muscle, hepatocyte, adipose tissue etc., and (5) genetics of gene expression. The microarray evidences of molecular basis of obesity and insulin resistance are presented here. Despite the limitations, microarray has potential clinical applications in finding new molecular targets for treatment of insulin resistance and classification of adipose tissue based on future risk of insulin resistance syndrome.

Consistency of predictive signature genes and classifiers generated using different microarray platforms

Microarray-based classifiers and associated signature genes generated from various platforms are abundantly reported in the literature; however, the utility of the classifiers and signature genes in cross-platform prediction applications remains largely uncertain. As part of the MicroArray Quality Control Phase II (MAQC-II) project, we show in this study 80-90% cross-platform prediction consistency using a large toxicogenomics data set by illustrating that: (1) the signature genes of a classifier generated from one platform can be directly applied to another platform to develop a predictive classifier; (2) a classifier developed using data generated from one platform can accurately predict samples that were profiled using a different platform. The results suggest the potential utility of using published signature genes in cross-platform applications and the possible adoption of the published classifiers for a variety of applications. The study reveals an opportunity for possible translation of biomarkers identified using microarrays to clinically validated non-array gene expression assays.

Practical application of toxicogenomics for profiling toxicant-induced biological perturbations

A systems-level understanding of molecular perturbations is crucial for evaluating chemical-induced toxicity risks appropriately, and for this purpose comprehensive gene expression analysis or toxicogenomics investigation is highly advantageous. The recent accumulation of toxicity-associated gene sets (toxicogenomic biomarkers), enrichment in public or commercial large-scale microarray database and availability of open-source software resources facilitate our utilization of the toxicogenomic data. However, toxicologists, who are usually not experts in computational sciences, tend to be overwhelmed by the gigantic amount of data. In this paper we present practical applications of toxicogenomics by utilizing biomarker gene sets and a simple scoring method by which overall gene set-level expression changes can be evaluated efficiently. Results from the gene set-level analysis are not only an easy interpretation of toxicological significance compared with individual gene-level profiling, but also are thought to be suitable for cross-platform or cross-institutional toxicogenomics data analysis. Enrichment in toxicogenomics databases, refinements of biomarker gene sets and scoring algorithms and the development of user-friendly integrative software will lead to better evaluation of toxicant-elicited biological perturbations.

A gene expression signature of RAS pathway dependence predicts response to PI3K and RAS pathway inhibitors and expands the population of RAS pathway activated tumors

Background

Hyperactivation of the Ras signaling pathway is a driver of many cancers, and RAS pathway activation can predict response to targeted therapies. Therefore, optimal methods for measuring Ras pathway activation are critical. The main focus of our work was to develop a gene expression signature that is predictive of RAS pathway dependence.

Methods

We used the coherent expression of RAS pathway-related genes across multiple datasets to derive a RAS pathway gene expression signature and generate RAS pathway activation scores in pre-clinical cancer models and human tumors. We then related this signature to KRAS mutation status and drug response data in pre-clinical and clinical datasets.

Results

The RAS signature score is predictive of KRAS mutation status in lung tumors and cell lines with high (> 90%) sensitivity but relatively low (50%) specificity due to samples that have apparent RAS pathway activation in the absence of a KRAS mutation. In lung and breast cancer cell line panels, the RAS pathway signature score correlates with pMEK and pERK expression, and predicts resistance to AKT inhibition and sensitivity to MEK inhibition within both KRAS mutant and KRAS wild-type groups. The RAS pathway signature is upregulated in breast cancer cell lines that have acquired resistance to AKT inhibition, and is downregulated by inhibition of MEK. In lung cancer cell lines knockdown of KRAS using siRNA demonstrates that the RAS pathway signature is a better measure of dependence on RAS compared to KRAS mutation status. In human tumors, the RAS pathway signature is elevated in ER negative breast tumors and lung adenocarcinomas, and predicts resistance to cetuximab in metastatic colorectal cancer.

Conclusions

These data demonstrate that the RAS pathway signature is superior to KRAS mutation status for the prediction of dependence on RAS signaling, can predict response to PI3K and RAS pathway inhibitors, and is likely to have the most clinical utility in lung and breast tumors.

Bimodal gene expression patterns in breast cancer

We identified a set of genes with an unexpected bimodal distribution among breast cancer patients in multiple studies. The property of bimodality seems to be common, as these genes were found on multiple microarray platforms and in studies with different end-points and patient cohorts. Bimodal genes tend to cluster into small groups of four to six genes with synchronised expression within the group (but not between the groups), which makes them good candidates for robust conditional descriptors. The groups tend to form concise network modules underlying their function in cancerogenesis of breast neoplasms.

Towards large-scale sample annotation in gene expression repositories

Background

Large repositories of biomedical research data are most useful to translational researchers if their data can be aggregated for efficient queries and analyses. However, inconsistent or non-existent annotations describing important sample details such as name of tissue or cell line, histopathological type, and subject characteristics like demographics, treatment, and survival are seldom present in data repositories, making it difficult to aggregate data.

Results

We created a flexible software tool that allows efficient annotation of samples using a controlled vocabulary, and report on its use for the annotation of over 12,500 samples.

Conclusion

While the amount of data is very large and seemingly poorly annotated, a lot of information is still within reach. Consistent tool-based re-annotation enables many new possibilities for large scale interpretation and analyses that would otherwise be impossible.

UVB radiation induces persistent activation of ribosome and oxidative phosphorylation pathways

Ultraviolet B (UVB) radiation has strong biological effects and modulates the expression of many genes. The major biological pathways affected by UVB radiation remain controversial. In this work, we used a loop-design microarray approach and applied rigorous statistical analyses to identify differentially regulated genes at 4, 8, 16 or 24 h after UVB irradiation. The most prominent biological categories in lists of differentially regulated gene sets were extracted by functional enrichment analysis. With this approach, we determined that genes participating in two prime cellular processes, the ribosome pathway and the oxidative phosphorylation pathway, were persistently activated after UVB irradiation. Mitochondrial activity assays confirmed increased activity for up to 24 h after UVB irradiation. These results suggest that the persistent activation of ribosome and oxidative phosphorylation pathways may have a key role in UVB-radiation-induced cellular responses. For the first time, the specific cellular pathways that respond to UVB radiation consistently and persistently can be delineated with confidence using a loop-design microarray approach and functional bioinformatics analysis. The results of this study offer further insight into UVB-radiation-induced stress responses.

Toxicogenomic biomarkers for liver toxicity

Toxicogenomics (TGx) is a widely used technique in the preclinical stage of drug development to investigate the molecular mechanisms of toxicity. A number of candidate TGx biomarkers have now been identified and are utilized for both assessing and predicting toxicities. Further accumulation of novel TGx biomarkers will lead to more efficient, appropriate and cost effective drug risk assessment, reinforcing the paradigm of the conventional toxicology system with a more profound understanding of the molecular mechanisms of drug-induced toxicity. In this paper, we overview some practical strategies as well as obstacles for identifying and utilizing TGx biomarkers based on microarray analysis. Since clinical hepatotoxicity is one of the major causes of drug development attrition, the liver has been the best documented target organ for TGx studies to date, and we therefore focused on information from liver TGx studies. In this review, we summarize the current resources in the literature in regard to TGx studies of the liver, from which toxicologists could extract potential TGx biomarker gene sets for better hepatotoxicity risk assessment.

Using differential gene expression to study Entamoeba histolytica pathogenesis

The release of the Entamoeba histolytica genome has facilitated the development of techniques to survey rapidly and to relate gene expression with biology. The association and potential contribution of differential gene expression to the life cycle and the virulence of this protozoan parasite of humans are reviewed here.

Novel insights into adipogenesis from omics data

Obesity, the excess accumulation of adipose tissue, is one of the most pressing health problems in both the Western world and in developing countries. Adipose tissue growth results from two processes: the increase in number of adipocytes (hyperplasia) that develop from precursor cells, and the growth of individual fat cells (hypertrophy) due to incorporation of triglycerides. Adipogenesis, the process of fat cell development, has been extensively studied using various cell and animal models. While these studies pointed out a number of key factors involved in adipogenesis, the list of molecular components is far from complete. The advance of high-throughput technologies has sparked many experimental studies aimed at the identification of novel molecular components regulating adipogenesis. This paper examines the results of recent studies on adipogenesis using high-throughput technologies. Specifically, it provides an overview of studies employing microarrays for gene expression profiling and studies using gel based and non-gel based proteomics as well as a chromatin immunoprecipitation followed by microarray analysis (ChIP-chip) or sequencing (ChIP-seq). Due to the maturity of the technology, the bulk of the available data was generated using microarrays. Therefore these data sets were not only reviewed but also underwent meta analysis. The review also shows that large-scale omics technologies in conjunction with sophisticated bioinformatics analyses can provide not only a list of novel players, but also a global view on biological processes and molecular networks. Finally, developing technologies and computational challenges associated with the data analyses are highlighted, and an outlook on the questions not previously addressed is provided.

SQUAT: A web tool to mine human, murine and avian SAGE data

Background

There is an increasing need in transcriptome research for gene expression data and pattern warehouses. It is of importance to integrate in these warehouses both raw transcriptomic data, as well as some properties encoded in these data, like local patterns.

Description

We have developed an application called SQUAT (SAGE Querying and Analysis Tools) which is available at: . This database gives access to both raw SAGE data and patterns mined from these data, for three species (human, mouse and chicken). This database allows to make simple queries like "In which biological situations is my favorite gene expressed?" as well as much more complex queries like: ≪what are the genes that are frequently co-over-expressed with my gene of interest in given biological situations?≫. Connections with external web databases enrich biological interpretations, and enable sophisticated queries. To illustrate the power of SQUAT, we show and analyze the results of three different queries, one of which led to a biological hypothesis that was experimentally validated.

Conclusion

SQUAT is a user-friendly information retrieval platform, which aims at bringing some of the state-of-the-art mining tools to biologists.

Meta-analysis of microarray results: challenges, opportunities, and recommendations for standardization

Microarray profiling of gene expression is a powerful tool for discovery, but the ability to manage and compare the resulting data can be problematic. Biological, experimental, and technical variations between studies of the same phenotype/phenomena create substantial differences in results. The application of conventional meta-analysis to raw microarray data is complicated by differences in the type of microarray used, gene nomenclatures, species, and analytical methods. An alternative approach to combining multiple microarray studies is to compare the published gene lists which result from the investigators' analyses of the raw data, as implemented in Lists of Lists Annotated (LOLA: www.lola.gwu.edu) and L2L (depts.washington.edu/l2l/). The present review considers both the potential value and the limitations of databasing and enabling the comparison of results from different microarray studies. Further, a major impediment to cross-study comparisons is the absence of a standard for reporting microarray study results. We propose a reporting standard: standard microarray results template (SMART), which will facilitate the integration of microarray studies.

Sensitive and robust gene expression changes in fish exposed to estrogen--a microarray approach

Background

Vitellogenin is a well established biomarker for estrogenic exposure in fish. However, effects on gonadal differentiation at concentrations of estrogen not sufficient to give rise to a measurable vitellogenin response suggest that more sensitive biomarkers would be useful. Induction of zona pellucida genes may be more sensitive but their specificities are not as clear. The objective of this study was to find additional sensitive and robust candidate biomarkers of estrogenic exposure.

Results

Hepatic mRNA expression profiles were characterized in juvenile rainbow trout exposed to a measured concentration of 0.87 and 10 ng ethinylestradiol/L using a salmonid cDNA microarray. The higher concentration was used to guide the subsequent identification of generally more subtle responses at the low concentration not sufficient to induce vitellogenin. A meta-analysis was performed with data from the present study and three similar microarray studies using different fish species and platforms. Within the generated list of presumably robust responses, several well-known estrogen-regulated genes were identified. Two genes, confirmed by quantitative RT-PCR (qPCR), fulfilled both the criteria of high sensitivity and robustness; the induction of the genes encoding zona pellucida protein 3 and a nucleoside diphosphate kinase (nm23).

Conclusion

The cross-species, cross-platform meta-analysis correctly identified several robust responses. This adds confidence to our approach used for identifying candidate biomarkers. Specifically, we propose that analyses of an nm23 gene together with zona pellucida genes may increase the possibilities to detect an exposure to low levels of estrogenic compounds in fish.

Review of the literature examining the correlation among DNA microarray technologies

DNA microarray technologies are used in a variety of biological disciplines. The diversity of platforms and analytical methods employed has raised concerns over the reliability, reproducibility and correlation of data produced across the different approaches. Initial investigations (years 2000-2003) found discrepancies in the gene expression measures produced by different microarray technologies. Increasing knowledge and control of the factors that result in poor correlation among the technologies has led to much higher levels of correlation among more recent publications (years 2004 to present). Here, we review the studies examining the correlation among microarray technologies. We find that with improvements in the technology (optimization and standardization of methods, including data analysis) and annotation, analysis across platforms yields highly correlated and reproducible results. We suggest several key factors that should be controlled in comparing across technologies, and are good microarray practice in general.

Microarray analysis of the AHR system: Tissue-specific flexibility in signal and target genes

Data mining published microarray experiments require that expression profiles are directly comparable. We performed linear global normalization on the data of 1967 Affymetrix U74av2 microarrays, i.e. the transcriptomes of >100 murine tissues or cell types. The mathematical transformation effectively nullifies inter-experimental or inter-laboratory differences between microarrays. The correctness of expression values was validated by quantitative RT-PCR. Using the database we analyze components of the aryl hydrocarbon receptor (AHR) signaling pathway in various tissues. We identified lineage and differentiation specific variant expression of AHR, ARNT, and HIF1alpha in the T-cell lineage and high expression of CYP1A1 in immature B cells and dendritic cells. Performing co-expression analysis we found unorthodox expression of the AHR in the absence of ARNT, particularly in stem cell populations, and can reject the hypothesis that ARNT2 takes over and is highly expressed when ARNT expression is low or absent. Furthermore the AHR shows no co-expression with any other transcript present on the chip. Analysis of differential gene expression under 308 conditions revealed 53 conditions under which the AHR is regulated, numerous conditions under which an intrinsic AHR action is modified as well as conditions activating the AHR even in the absence of known AHR ligands. Thus meta-analysis of published expression profiles is a powerful tool to gain novel insights into known and unknown systems.

Statistical analysis of an RNA titration series evaluates microarray precision and sensitivity on a whole-array basis

BACKGROUND

Concerns are often raised about the accuracy of microarray technologies and the degree of cross-platform agreement, but there are yet no methods which can unambiguously evaluate precision and sensitivity for these technologies on a whole-array basis.

RESULTS

A methodology is described for evaluating the precision and sensitivity of whole-genome gene expression technologies such as microarrays. The method consists of an easy-to-construct titration series of RNA samples and an associated statistical analysis using non-linear regression. The method evaluates the precision and responsiveness of each microarray platform on a whole-array basis, i.e., using all the probes, without the need to match probes across platforms. An experiment is conducted to assess and compare four widely used microarray platforms. All four platforms are shown to have satisfactory precision but the commercial platforms are superior for resolving differential expression for genes at lower expression levels. The effective precision of the two-color platforms is improved by allowing for probe-specific dye-effects in the statistical model. The methodology is used to compare three data extraction algorithms for the Affymetrix platforms, demonstrating poor performance for the commonly used proprietary algorithm relative to the other algorithms. For probes which can be matched across platforms, the cross-platform variability is decomposed into within-platform and between-platform components, showing that platform disagreement is almost entirely systematic rather than due to measurement variability.

CONCLUSION

The results demonstrate good precision and sensitivity for all the platforms, but highlight the need for improved probe annotation. They quantify the extent to which cross-platform measures can be expected to be less accurate than within-platform comparisons for predicting disease progression or outcome.