Tapan Mehta

Repeatability of published microarray gene expression analyses

Given the complexity of microarray-based gene expression studies, guidelines encourage transparent design and public data availability. Several journals require public data deposition and several public databases exist. However, not all data are publicly available, and even when available, it is unknown whether the published results are reproducible by independent scientists. Here we evaluated the replication of data analyses in 18 articles on microarray-based gene expression profiling published in Nature Genetics in 2005–2006. One table or figure from each article was independently evaluated by two teams of analysts. We reproduced two analyses in principle and six partially or with some discrepancies; ten could not be reproduced. The main reason for failure to reproduce was data unavailability, and discrepancies were mostly due to incomplete data annotation or specification of data processing and analysis. Repeatability of published microarray studies is apparently limited. More strict publication rules enforcing public data availability and explicit description of data processing and analysis should be considered.

Myths, Presumptions, and Facts about Obesity

BACKGROUND Many beliefs about obesity persist in the absence of supporting scientific evidence (presumptions); some persist despite contradicting evidence (myths). The promulgation of unsupported beliefs may yield poorly informed policy decisions, inaccurate clinical and public health recommendations, and an unproductive allocation of research resources and may divert attention away from useful, evidence-based information. METHODS Using Internet searches of popular media and scientific literature, we identified, reviewed, and classified obesity-related myths and presumptions. We also examined facts that are well supported by evidence, with an emphasis on those that have practical implications for public health, policy, or clinical recommendations. RESULTS We identified seven obesity-related myths concerning the effects of small sustained increases in energy intake or expenditure, establishment of realistic goals for weight loss, rapid weight loss, weight-loss readiness, physical-education classes, breast-feeding, and energy expended during sexual activity. We also identified six presumptions about the purported effects of regularly eating breakfast, early childhood experiences, eating fruits and vegetables, weight cycling, snacking, and the built (i.e., human-made) environment. Finally, we identified nine evidence-supported facts that are relevant for the formulation of sound public health, policy, or clinical recommendations. CONCLUSIONS False and scientifically unsupported beliefs about obesity are pervasive in both scientific literature and the popular press. (Funded by the National Institutes of Health.).

Transcriptional coordination of the metabolic network in Arabidopsis

Patterns of coexpression can reveal networks of functionally related genes and provide deeper understanding of processes requiring multiple gene products. We performed an analysis of coexpression networks for 1,330 genes from the AraCyc database of metabolic pathways in Arabidopsis (Arabidopsis thaliana). We found that genes associated with the same metabolic pathway are, on average, more highly coexpressed than genes from different pathways. Positively coexpressed genes within the same pathway tend to cluster close together in the pathway structure, while negatively correlated genes typically occupy more distant positions. The distribution of coexpression links per gene is highly skewed, with a small but significant number of genes having numerous coexpression partners but most having fewer than 10. Genes with multiple connections (hubs) tend to be single-copy genes, while genes with multiple paralogs are coexpressed with fewer genes, on average, than single-copy genes, suggesting that the network expands through gene duplication, followed by weakening of coexpression links involving duplicate nodes. Using a network-analysis algorithm based on coexpression with multiple pathway members (pathway-level coexpression), we identified and prioritized novel candidate pathway members, regulators, and cross pathway transcriptional control points for over 140 metabolic pathways. To facilitate exploration and analysis of the results, we provide a Web site (http://www.transvar.org/at_coexpress/analysis/web) listing analyzed pathways with links to regression and pathway-level coexpression results. These methods and results will aid in the prioritization of candidates for genetic analysis of metabolism in plants and contribute to the improvement of functional annotation of the Arabidopsis genome.

CressExpress: A Tool For Large-Scale Mining of Expression Data from Arabidopsis

CressExpress is a user-friendly, online, coexpression analysis tool for Arabidopsis (Arabidopsis thaliana) microarray expression data that computes patterns of correlated expression between user-entered query genes and the rest of the genes in the genome. Unlike other coexpression tools, CressExpress allows characterization of tissue-specific coexpression networks through user-driven filtering of input data based on sample tissue type. CressExpress also performs pathway-level coexpression analysis on each set of query genes, identifying and ranking genes based on their common connections with two or more query genes. This allows identification of novel candidates for involvement in common processes and functions represented by the query group. Users launch experiments using an easy-to-use Web-based interface and then receive the full complement of results, along with a record of tool settings and parameters, via an e-mail link to the CressExpress Web site. Data sets featured in CressExpress are strictly versioned and include expression data from MAS5, GCRMA, and RMA array processing algorithms. To demonstrate applications for CressExpress, we present coexpression analyses of cellulose synthase genes, indolic glucosinolate biosynthesis, and flowering. We show that subselecting sample types produces a richer network for genes involved in flowering in Arabidopsis. CressExpress provides direct access to expression values via an easy-to-use URL-based Web service, allowing users to determine quickly if their query genes are coexpressed with each other and likely to yield informative pathway-level coexpression results. The tool is available at http://www.cressexpress.org.

Sources of variation in Affymetrix microarray experiments

BackgroundA typical microarray experiment has many sources of variation which can be attributed to biological and technical causes. Identifying sources of variation and assessing their magnitude, among other factors, are important for optimal experimental design. The objectives of this study were: (1) to estimate relative magnitudes of different sources of variation and (2) to evaluate agreement between biological and technical replicates.ResultsWe performed a microarray experiment using a total of 24 Affymetrix GeneChip® arrays. The study included 4th mammary gland samples from eight 21-day-old Sprague Dawley CD female rats exposed to genistein (soy isoflavone). RNA samples from each rat were split to assess variation arising at labeling and hybridization steps. A general linear model was used to estimate variance components. Pearson correlations were computed to evaluate agreement between technical and biological replicates.ConclusionThe greatest source of variation was biological variation, followed by residual error, and finally variation due to labeling when *.cel files were processed with dChip and RMA image processing algorithms. When MAS 5.0 or GCRMA-EB were used, the greatest source of variation was residual error, followed by biology and labeling. Correlations between technical replicates were consistently higher than between biological replicates.

R/qtlbim: QTL with Bayesian Interval Mapping in experimental crosses

UNLABELLED R/qtlbim is an extensible, interactive environment for the Bayesian Interval Mapping of QTL, built on top of R/qtl (Broman et al., 2003), providing Bayesian analysis of multiple interacting quantitative trait loci (QTL) models for continuous, binary and ordinal traits in experimental crosses. It includes several efficient Markov chain Monte Carlo (MCMC) algorithms for evaluating the posterior of genetic architectures, i.e. the number and locations of QTL, their main and epistatic effects and gene-environment interactions. R/qtlbim provides extensive informative graphical and numerical summaries, and model selection and convergence diagnostics of the MCMC output, illustrated through the vignette, example and demo capabilities of R (R Development Core Team 2006). AVAILABILITY The package is freely available from cran.r-project.org.

Use of self-reported height and weight biases the body mass index-mortality association

Background:Many large-scale epidemiological data sources used to evaluate the body mass index (BMI: kg/m2) mortality association have relied on BMI derived from self-reported height and weight. Although measured BMI (BMIM) and self-reported BMI (BMISR) correlate highly, self-reports are systematically biased.Objective:To rigorously examine how self-reporting bias influences the association between BMI and mortality rate.Subjects:Samples representing the US non-institutionalized civilian population.Design and Methods:National Health and Nutrition Examination Survey data (NHANES II: 1976–80; NHANES III: 1988–94) contain BMIM and BMISR. We applied Cox regression to estimate mortality hazard ratios (HRs) for BMIM and BMISR categories, respectively, and compared results. We similarly analyzed subgroups of ostensibly healthy never-smokers.Results:Misclassification by BMISR among the underweight and obesity ranged from 30–40% despite high correlations between BMIM and BMISR (r>0.9). The reporting bias was moderately correlated with BMIM (r>0.35), but not BMISR (r<0.15). Analyses using BMISR failed to detect six of eight significant mortality HRs detected by BMIM. Significantly biased HRs were detected in the NHANES II full data set (χ2=12.49; P=0.01) and healthy subgroup (χ2=9.93; P=0.04), but not in the NHANES III full data set (χ2=5.63; P=0.23) or healthy subgroup (χ2=1.52; P=0.82).Conclusions:BMISR should not be treated as interchangeable with BMIM in BMI mortality analyses. Bias and inconsistency introduced by using BMISR in place of BMIM in BMI mortality estimation and hypothesis tests may account for important discrepancies in published findings.

Integrative genomic analyses of neurofibromatosis tumours identify SOX9 as a biomarker and survival gene

Understanding the biological pathways critical for common neurofibromatosis type 1 (NF1) peripheral nerve tumours is essential, as there is a lack of tumour biomarkers, prognostic factors and therapeutics. We used gene expression profiling to define transcriptional changes between primary normal Schwann cells (n = 10), NF1‐derived primary benign neurofibroma Schwann cells (NFSCs) (n = 22), malignant peripheral nerve sheath tumour (MPNST) cell lines (n = 13), benign neurofibromas (NF) (n = 26) and MPNST (n = 6). Dermal and plexiform NFs were indistinguishable. A prominent theme in the analysis was aberrant differentiation. NFs repressed gene programs normally active in Schwann cell precursors and immature Schwann cells. MPNST signatures strongly differed; genes up‐regulated in sarcomas were significantly enriched for genes activated in neural crest cells. We validated the differential expression of 82 genes including the neural crest transcription factor SOX9 and SOX9 predicted targets. SOX9 immunoreactivity was robust in NF and MPSNT tissue sections and targeting SOX9 – strongly expressed in NF1‐related tumours – caused MPNST cell death. SOX9 is a biomarker of NF and MPNST, and possibly a therapeutic target in NF1.

An Efficient Bayesian Model Selection Approach for Interacting Quantitative Trait Loci Models With Many Effects

We extend our Bayesian model selection framework for mapping epistatic QTL in experimental crosses to include environmental effects and gene–environment interactions. We propose a new, fast Markov chain Monte Carlo algorithm to explore the posterior distribution of unknowns. In addition, we take advantage of any prior knowledge about genetic architecture to increase posterior probability on more probable models. These enhancements have significant computational advantages in models with many effects. We illustrate the proposed method by detecting new epistatic and gene–sex interactions for obesity-related traits in two real data sets of mice. Our method has been implemented in the freely available package R/qtlbim (http://www.qtlbim.org) to facilitate the general usage of the Bayesian methodology for genomewide interacting QTL analysis.

Correlation of microRNA levels during hypoxia with predicted target mRNAs through genome-wide microarray analysis

BackgroundLow levels of oxygen in tissues, seen in situations such as chronic lung disease, necrotic tumors, and high altitude exposures, initiate a signaling pathway that results in active transcription of genes possessing a hypoxia response element (HRE). The aim of this study was to investigate whether a change in miRNA expression following hypoxia could account for changes in the cellular transcriptome based on currently available miRNA target prediction tools.MethodsTo identify changes induced by hypoxia, we conducted mRNA- and miRNA-array-based experiments in HT29 cells, and performed comparative analysis of the resulting data sets based on multiple target prediction algorithms. To date, few studies have investigated an environmental perturbation for effects on genome-wide miRNA levels, or their consequent influence on mRNA output.ResultsComparison of miRNAs with predicted mRNA targets indicated a lower level of concordance than expected. We did, however, find preliminary evidence of combinatorial regulation of mRNA expression by miRNA.ConclusionTarget prediction programs and expression profiling techniques do not yet adequately represent the complexity of miRNA-mediated gene repression, and new methods may be required to better elucidate these pathways. Our data suggest the physiologic impact of miRNAs on cellular transcription results from a multifaceted network of miRNA and mRNA relationships, working together in an interconnected system and in context of hundreds of RNA species. The methods described here for comparative analysis of cellular miRNA and mRNA will be useful for understanding genome wide regulatory responsiveness and refining miRNA predictive algorithms.