A semi-parametric statistical model for integrating gene expression profiles across different platforms

Determining differentially expressed genes (DEGs) between biological samples is the key to understand how genotype gives rise to phenotype. RNA-seq and microarray are two main technologies for profiling gene expression levels. However, considerable discrepancy has been found between DEGs detected using the two technologies. Integration data across these two platforms has the potential to improve the power and reliability of DEG detection. We propose a rank-based semi-parametric model to determine DEGs using information across different sources and apply it to the integration of RNA-seq and microarray data. By incorporating both the significance of differential expression and the consistency across platforms, our method effectively detects DEGs with moderate but consistent signals. We demonstrate the effectiveness of our method using simulation studies, MAQC/SEQC data and a synthetic microRNA dataset. Our integration method is not only robust to noise and heterogeneity in the data, but also adaptive to the structure of data. In our simulations and real data studies, our approach shows a higher discriminate power and identifies more biologically relevant DEGs than eBayes, DEseq and some commonly used meta-analysis methods.     

Evolutionary rate and gene expression across different brain regions

Background

The evolutionary rate of a protein is a basic measure of evolution at the molecular level. Previous studies have shown that genes expressed in the brain have significantly lower evolutionary rates than those expressed in somatic tissues.

Results

We study the evolutionary rates of genes expressed in 21 different human brain regions. We find that genes highly expressed in the more recent cortical regions of the brain have lower evolutionary rates than genes highly expressed in subcortical regions. This may partially result from the observation that genes that are highly expressed in cortical regions tend to be highly expressed in subcortical regions, and thus their evolution faces a richer set of functional constraints. The frequency of mammal-specific and primate-specific genes is higher in the highly expressed gene sets of subcortical brain regions than in those of cortical brain regions. The basic inverse correlation between evolutionary rate and gene expression is significantly stronger in brain versus nonbrain tissues, and in cortical versus subcortical regions. Extending upon this cortical/subcortical trend, this inverse correlation is generally more marked for tissues that are located higher along the cranial vertical axis during development, giving rise to the possibility that these tissues are also more evolutionarily recent.

Conclusions

We find that cortically expressed genes are more conserved than subcortical ones, and that gene expression levels exert stronger constraints on sequence evolution in cortical versus subcortical regions. Taken together, these findings suggest that cortically expressed genes are under stronger selective pressure than subcortically expressed genes.     

A comparison of Affymetrix gene expression arrays

Background  

Affymetrix GeneChips™ are an important tool in many facets of biological research. Recently, notable design changes to the chips have been made. In this study, we use publicly available data from Affymetrix to gauge the performance of three human gene expression arrays: Human Genome U133 Plus 2.0 (U133), Human Exon 1.0 ST (HuEx) and Human Gene 1.0 ST (HuGene).     

Quantifying engineered nanomaterial toxicity: comparison of common cytotoxicity and gene expression measurements

Background

When evaluating the toxicity of engineered nanomaterials (ENMS) it is important to use multiple bioassays based on different mechanisms of action. In this regard we evaluated the use of gene expression and common cytotoxicity measurements using as test materials, two selected nanoparticles with known differences in toxicity, 5 nm mercaptoundecanoic acid (MUA)-capped InP and CdSe quantum dots (QDs). We tested the effects of these QDs at concentrations ranging from 0.5 to 160 µg/mL on cultured normal human bronchial epithelial (NHBE) cells using four common cytotoxicity assays: the dichlorofluorescein assay for reactive oxygen species (ROS), the lactate dehydrogenase assay for membrane viability (LDH), the mitochondrial dehydrogenase assay for mitochondrial function, and the Comet assay for DNA strand breaks.

Results

The cytotoxicity assays showed similar trends when exposed to nanoparticles for 24 h at 80 µg/mL with a threefold increase in ROS with exposure to CdSe QDs compared to an insignificant change in ROS levels after exposure to InP QDs, a twofold increase in the LDH necrosis assay in NHBE cells with exposure to CdSe QDs compared to a 50% decrease for InP QDs, a 60% decrease in the mitochondrial function assay upon exposure to CdSe QDs compared to a minimal increase in the case of InP and significant DNA strand breaks after exposure to CdSe QDs compared to no significant DNA strand breaks with InP. High-throughput quantitative real-time polymerase chain reaction (qRT-PCR) data for cells exposed for 6 h at a concentration of 80 µg/mL were consistent with the cytotoxicity assays showing major differences in DNA damage, DNA repair and mitochondrial function gene regulatory responses to the CdSe and InP QDs. The BRCA2, CYP1A1, CYP1B1, CDK1, SFN and VEGFA genes were observed to be upregulated specifically from increased CdSe exposure and suggests their possible utility as biomarkers for toxicity.

Conclusions

This study can serve as a model for comparing traditional cytotoxicity assays and gene expression measurements and to determine candidate biomarkers for assessing the biocompatibility of ENMs.

     

Analysis of matched mRNA measurements from two different microarray technologies

MOTVIATION: The existence of several technologies for measuring gene expression makes the question of cross-technology agreement of measurements an important issue. Cross-platform utilization of data from different technologies has the potential to reduce the need to duplicate experiments but requires corresponding measurements to be comparable. METHODS: A comparison of mRNA measurements of 2895 sequence-matched genes in 56 cell lines from the standard panel of 60 cancer cell lines from the National Cancer Institute (NCI 60) was carried out by calculating correlation between matched measurements and calculating concordance between cluster from two high-throughput DNA microarray technologies, Stanford type cDNA microarrays and Affymetrix oligonucleotide microarrays. RESULTS: In general, corresponding measurements from the two platforms showed poor correlation. Clusters of genes and cell lines were discordant between the two technologies, suggesting that relative intra-technology relationships were not preserved. GC-content, sequence length, average signal intensity, and an estimator of cross-hybridization were found to be associated with the degree of correlation. This suggests gene-specific, or more correctly probe-specific, factors influencing measurements differently in the two platforms, implying a poor prognosis for a broad utilization of gene expression measurements across platforms.     

Classification across gene expression microarray studies

Background  

The increasing number of gene expression microarray studies represents an important resource in biomedical research. As a result, gene expression based diagnosis has entered clinical practice for patient stratification in breast cancer. However, the integration and combined analysis of microarray studies remains still a challenge. We assessed the potential benefit of data integration on the classification accuracy and systematically evaluated the generalization performance of selected methods on four breast cancer studies comprising almost 1000 independent samples. To this end, we introduced an evaluation framework which aims to establish good statistical practice and a graphical way to monitor differences. The classification goal was to correctly predict estrogen receptor status (negative/positive) and histological grade (low/high) of each tumor sample in an independent study which was not used for the training. For the classification we chose support vector machines (SVM), predictive analysis of microarrays (PAM), random forest (RF) and k-top scoring pairs (kTSP). Guided by considerations relevant for classification across studies we developed a generalization of kTSP which we evaluated in addition. Our derived version (DV) aims to improve the robustness of the intrinsic invariance of kTSP with respect to technologies and preprocessing.     

Transport measurements across Caco-2 monolayers of different organic and inorganic selenium

The transport and uptake of the most common Se compounds, selenate (SeO ₄ ²⁻ ), selenite (SeO ₃ ²⁻ ), selenomethionine, and selenocystine, were investigated using confluent monolayers of Caco-2 cells, a human carcinoma cell line. Comparative measurements were performed in the absorptive (apical to basolateral side) and exsorptive (basolateral to apical side) directions. Apparent permeability coefficients (P _app), calculated from transport experiments in the absorptive direction, showed increasing values in the following rank order: about 1×10⁻⁶ cm/s ≤ mannitol ≤ SeO ₃ ²⁻ ≤ selenocystine < selenomethionine < SeO ₄ ²⁻ ≤ about 16×10⁻⁶ cm/s. The ratios of the P _app measured in the absorptive versus exsorptive directions indicated that only the organic forms presented a net polarized transport (P _app ratio ≫1), suggesting the presence of a transcellular pathway. No significant excretion was observed. The transport of selenomethionine was inhibited by its sulfur analog, methionine, suggesting a common transport mechanism. In contrast, an inhibition of the transport of selenocystine by cysteine was not observed. From the two substrates tested, sulfate and thiosulfate, only thiosulfate inhibited the transport of SeO ₄ ²⁻ . This effect was also observed for SeO ₃ ²⁻ (i.e., was unspecific), which questioned the assertion of a common transport for sulfate and SeO ₄ ²⁻ and may confirm the paracellular pathway of SeO ₄ ²⁻ suggested by the P _app ratio of about 1. The addition of glutathione (GSH) in large excess had no consequence on the passage of SeO ₃ ²⁻ but strongly increased the uptake (about fourfold). The liquid chromatography — mass spectrometry (LC-MS) data showed that, in the ionic condition of incubation medium, GSH promptly reduced SeO ₃ ²⁻ (≤2 min) in its elemental form Se⁰, which cannot ascribe to selenodiglutathione a direct role in the effect of GSH.     

Evaluation of gene expression measurements from commercial microarray platforms

Multiple commercial microarrays for measuring genome-wide gene expression levels are currently available, including oligonucleotide and cDNA, single- and two-channel formats. This study reports on the results of gene expression measurements generated from identical RNA preparations that were obtained using three commercially available microarray platforms. RNA was collected from PANC-1 cells grown in serum-rich medium and at 24 h following the removal of serum. Three biological replicates were prepared for each condition, and three experimental replicates were produced for the first biological replicate. RNA was labeled and hybridized to microarrays from three major suppliers according to manufacturers’ protocols, and gene expression measurements were obtained using each platform’s standard software. For each platform, gene targets from a subset of 2009 common genes were compared. Correlations in gene expression levels and comparisons for significant gene expression changes in this subset were calculated, and showed considerable divergence across the different platforms, suggesting the need for establishing industrial manufacturing standards, and further independent and thorough validation of the technology.     

A systematic comparison and evaluation of biclustering methods for gene expression data

MOTIVATION: In recent years, there have been various efforts to overcome the limitations of standard clustering approaches for the analysis of gene expression data by grouping genes and samples simultaneously. The underlying concept, which is often referred to as biclustering, allows to identify sets of genes sharing compatible expression patterns across subsets of samples, and its usefulness has been demonstrated for different organisms and datasets. Several biclustering methods have been proposed in the literature; however, it is not clear how the different techniques compare with each other with respect to the biological relevance of the clusters as well as with other characteristics such as robustness and sensitivity to noise. Accordingly, no guidelines concerning the choice of the biclustering method are currently available. RESULTS: First, this paper provides a methodology for comparing and validating biclustering methods that includes a simple binary reference model. Although this model captures the essential features of most biclustering approaches, it is still simple enough to exactly determine all optimal groupings; to this end, we propose a fast divide-and-conquer algorithm (Bimax). Second, we evaluate the performance of five salient biclustering algorithms together with the reference model and a hierarchical clustering method on various synthetic and real datasets for Saccharomyces cerevisiae and Arabidopsis thaliana. The comparison reveals that (1) biclustering in general has advantages over a conventional hierarchical clustering approach, (2) there are considerable performance differences between the tested methods and (3) already the simple reference model delivers relevant patterns within all considered settings.     

Indirect measurements of differential gene expression with cDNA microarrays

The use of universal RNA reference sets is an increasingly common approach to molecular classification studies with cDNA microarrays. Here we evaluated the reliability of indirect measurements of fluorescence ratios with a common RNA reference as a means of identifying differentially expressed genes. Comparisons of direct and indirect measures of differential gene expression showed a strong overall correlation in fluorescence ratio measurements but also a high degree of false positives in our indirect measurements. These results indicated that the application of more stringent ratio filters may be required when assessing differential gene expression utilizing a common RNA reference in classification studies.     

A comparison of probe-level and probeset models for small-sample gene expression data

Background  

Statistical methods to tentatively identify differentially expressed genes in microarray studies typically assume larger sample sizes than are practical or even possible in some settings.     

cRNA target preparation for microarrays: comparison of gene expression profiles generated with different amplification procedures

Microarray technology has become a standard tool for generation of gene expression profiles to explore human disease processes. Being able to start from minute amounts of RNA extends the fields of application to core needle biopsies, laser capture microdissected cells, and flow-sorted cells. Several RNA amplification methods have been developed, but no extensive comparability and concordance studies of gene expression profiles are available. Different amplification methods may produce differences in gene expression patterns. Therefore, we compared profiles processed by a standard microarray protocol with three different types of RNA amplification: (i) two rounds of linear target amplification, (ii) random amplification, and (iii) amplification based on a template switching mechanism. The latter two methods accomplish target amplification in a nonlinear way using PCR technology. Starting from as little as 50 ng of total RNA, the yield of labeled cRNA was sufficient for hybridization to Affymetrix HG-U133A GeneChip array using the respective methods. Replicate experiments were highly reproducible for each method. In comparison with the standard protocol, all three approaches are less sensitive and introduced a minor but clearly detectable bias of the detection call. In conclusion, the three amplification protocols used are applicable for GeneChip analysis of small tissue samples.