Improving identification of differentially expressed genes by integrative analysis of Affymetrix and Illumina arrays

Together with the widely used Affymetrix microarrays, the recently introduced Illumina platform has become a cost-effective alternative for genome-wide studies. To efficiently use data from both array platforms, there is a pressing need for methods that allow systematic integration of multiple datasets, especially when the number of samples is small. To address these needs, we introduce a meta-analytic procedure for combining Affymetrix and Illumina data in the context of detecting differentially expressed genes between the platforms. We first investigate the effect of different expression change estimation procedures within the platforms on the agreement of the most differentially expressed genes. Using the best estimation methods, we then show the benefits of the integrative analysis in producing reproducible results across bootstrap samples. In particular, we demonstrate its biological relevance in identifying small but consistent changes during T helper 2 cell differentiation.     

Evaluating gene expression in C57BL/6J and DBA/2J mouse striatum using RNA-Seq and microarrays

C57BL/6J (B6) and DBA/2J (D2) are two of the most commonly used inbred mouse strains in neuroscience research. However, the only currently available mouse genome is based entirely on the B6 strain sequence. Subsequently, oligonucleotide microarray probes are based solely on this B6 reference sequence, making their application for gene expression profiling comparisons across mouse strains dubious due to their allelic sequence differences, including single nucleotide polymorphisms (SNPs). The emergence of next-generation sequencing (NGS) and the RNA-Seq application provides a clear alternative to oligonucleotide arrays for detecting differential gene expression without the problems inherent to hybridization-based technologies. Using RNA-Seq, an average of 22 million short sequencing reads were generated per sample for 21 samples (10 B6 and 11 D2), and these reads were aligned to the mouse reference genome, allowing 16,183 Ensembl genes to be queried in striatum for both strains. To determine differential expression, 'digital mRNA counting' is applied based on reads that map to exons. The current study compares RNA-Seq (Illumina GA IIx) with two microarray platforms (Illumina MouseRef-8 v2.0 and Affymetrix MOE 430 2.0) to detect differential striatal gene expression between the B6 and D2 inbred mouse strains. We show that by using stringent data processing requirements differential expression as determined by RNA-Seq is concordant with both the Affymetrix and Illumina platforms in more instances than it is concordant with only a single platform, and that instances of discordance with respect to direction of fold change were rare. Finally, we show that additional information is gained from RNA-Seq compared to hybridization-based techniques as RNA-Seq detects more genes than either microarray platform. The majority of genes differentially expressed in RNA-Seq were only detected as present in RNA-Seq, which is important for studies with smaller effect sizes where the sensitivity of hybridization-based techniques could bias interpretation.     

Gene Expression Ratios Lead to Accurate and Translatable Predictors of DR5 Agonism across Multiple Tumor Lineages

Death Receptor 5 (DR5) agonists demonstrate anti-tumor activity in preclinical models but have yet to demonstrate robust clinical responses. A key limitation may be the lack of patient selection strategies to identify those most likely to respond to treatment. To overcome this limitation, we screened a DR5 agonist Nanobody across >600 cell lines representing 21 tumor lineages and assessed molecular features associated with response. High expression of DR5 and Casp8 were significantly associated with sensitivity, but their expression thresholds were difficult to translate due to low dynamic ranges. To address the translational challenge of establishing thresholds of gene expression, we developed a classifier based on ratios of genes that predicted response across lineages. The ratio classifier outperformed the DR5+Casp8 classifier, as well as standard approaches for feature selection and classification using genes, instead of ratios. This classifier was independently validated using 11 primary patient-derived pancreatic xenograft models showing perfect predictions as well as a striking linearity between prediction probability and anti-tumor response. A network analysis of the genes in the ratio classifier captured important biological relationships mediating drug response, specifically identifying key positive and negative regulators of DR5 mediated apoptosis, including DR5, CASP8, BID, cFLIP, XIAP and PEA15. Importantly, the ratio classifier shows translatability across gene expression platforms (from Affymetrix microarrays to RNA-seq) and across model systems (in vitro to in vivo). Our approach of using gene expression ratios presents a robust and novel method for constructing translatable biomarkers of compound response, which can also probe the underlying biology of treatment response.     

A comparison of cDNA, oligonucleotide, and Affymetrix GeneChip gene expression microarray platforms.

We have conducted a study to compare the variability in measured gene expression levels associated with three types of microarray platforms. Total RNA samples were obtained from liver tissue of four male mice, two each from inbred strains A/J and C57BL/6J. The same four samples were assayed on Affymetrix Mouse Genome Expression Set 430 GeneChips (MOE430A and MOE430B), spotted cDNA microarrays, and spotted oligonucleotide microarrays using eight arrays of each type. Variances associated with measurement error were observed to be comparable across all microarray platforms. The MOE430A GeneChips and cDNA arrays had higher precision across technical replicates than the MOE430B GeneChips and oligonucleotide arrays. The Affymetrix platform showed the greatest range in the magnitude of expression levels followed by the oligonucleotide arrays. We observed good concordance in both estimated expression level and statistical significance of common genes between the Affymetrix MOE430A GeneChip and the oligonucleotide arrays. Despite their apparently high precision, cDNA arrays showed poor concordance with other platforms.     

RefSeq Refinements of UniGene-Based Gene Matching Improve the Correlation of Expression Measurements Between Two Microarray Platforms

Matching genes across microarray platforms is a critical step in meta-analysis. Standard practice uses UniGene to match genes. Numerous studies have found poor correlations between platforms when using UniGene matching.We profiled samples from 33 breast cancer patients on two different microarray platforms (Affymetrix and cDNA) and investigated gene matching. Our results confirmed that UniGene-based matching led to poor correlations of gene expression between platforms. Using RefSeq, a database maintained by the National Center for Biotechnology Information (NCBI), we developed and implemented a new method to refine gene matching. We found that the correlations between gene expression measurements were substantially higher after the RefSeq matching. Our approach differs from previously reported sequence-matching approaches and retains useful expression measurements. It is a sensible approach for matching probes across platforms.We conclude that UniGene alone is insufficient to match genes across platforms. Refined matching based on RefSeq significantly improves the quality of matches.     

Independence and reproducibility across microarray platforms

Microarrays have been widely used for the analysis of gene expression, but the issue of reproducibility across platforms has yet to be fully resolved. To address this apparent problem, we compared gene expression between two microarray platforms: the short oligonucleotide Affymetrix Mouse Genome 430 2.0 GeneChip and a spotted cDNA array using a mouse model of angiotensin II-induced hypertension. RNA extracted from treated mice was analyzed using Affymetrix and cDNA platforms and then by quantitative RT-PCR (qRT-PCR) for validation of specific genes. For the 11,710 genes present on both arrays, we assessed the relative impact of experimental treatment and platform on measured expression and found that biological treatment had a far greater impact on measured expression than did platform for more than 90% of genes, a result validated by qRT-PCR. In the small number of cases in which platforms yielded discrepant results, qRT-PCR generally did not confirm either set of data, suggesting that sequence-specific effects may make expression predictions difficult to make using any technique.     

Evaluation of Microarray Preprocessing Algorithms Based on Concordance with RT-PCR in Clinical Samples

Background

Several preprocessing algorithms for Affymetrix gene expression microarrays have been developed, and their performance on spike-in data sets has been evaluated previously. However, a comprehensive comparison of preprocessing algorithms on samples taken under research conditions has not been performed.

Methodology/Principal Findings

We used TaqMan RT-PCR arrays as a reference to evaluate the accuracy of expression values from Affymetrix microarrays in two experimental data sets: one comprising 84 genes in 36 colon biopsies, and the other comprising 75 genes in 29 cancer cell lines. We evaluated consistency using the Pearson correlation between measurements obtained on the two platforms. Also, we introduce the log-ratio discrepancy as a more relevant measure of discordance between gene expression platforms. Of nine preprocessing algorithms tested, PLIER+16 produced expression values that were most consistent with RT-PCR measurements, although the difference in performance between most of the algorithms was not statistically significant.

Conclusions/Significance

Our results support the choice of PLIER+16 for the preprocessing of clinical Affymetrix microarray data. However, other algorithms performed similarly and are probably also good choices.     

Preprocessing of gene expression data by optimally robust estimators

Background  

The preprocessing of gene expression data obtained from several platforms routinely includes the aggregation of multiple raw signal intensities to one expression value. Examples are the computation of a single expression measure based on the perfect match (PM) and mismatch (MM) probes for the Affymetrix technology, the summarization of bead level values to bead summary values for the Illumina technology or the aggregation of replicated measurements in the case of other technologies including real-time quantitative polymerase chain reaction (RT-qPCR) platforms. The summarization of technical replicates is also performed in other "-omics" disciplines like proteomics or metabolomics.     

An integrated cross-platform prognosis study on neuroblastoma patients

There have been several reports about the potential for predicting prognosis of neuroblastoma patients using microarray gene expression profiling of the tumors. However these studies have revealed an apparent diversity in the identity of the genes in their predictive signatures. To test the contribution of the platform to this discrepancy we applied the z-scoring method to minimize the impact of platform and combine gene expression profiles of neuroblastoma (NB) tumors from two different platforms, cDNA and Affymetrix. A total of 12442 genes were common to both cDNA and Affymetrix arrays in our data set. Two-way ANOVA analysis was applied to the combined data set for assessing the relative effect of prognosis and platform on gene expression. We found that 26.6% (3307) of the genes had significant impact on survival. There was no significant impact of microarray platform on expression after application of z-scoring standardization procedure. Artificial neural network (ANN) analysis of the combined data set in a leave-one-out prediction strategy correctly predicted the outcome for 90% of the samples. Hierarchical clustering analysis using the top-ranked 160 genes showed the great separation of two clusters, and the majority of matched samples from the different platforms were clustered next to each other. The ANN classifier trained with our combined cross-platform data for these 160 genes could predict the prognosis of 102 independent test samples with 71% accuracy. Furthermore it correctly predicted the outcome for 85/102 (83%) NB patients through the leave-one-out cross-validation approach. Our study showed that gene expression studies performed in different platforms could be integrated for prognosis analysis after removing variation resulting from different platforms.     

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.     

Modeling nonlinearity in dilution design microarray data

MOTIVATION: Dilution design (Mixed tissue RNA) has been utilized by some researchers to evaluate and assess the performance of multiple microarray platforms. Current microarray data analysis approaches assume that the quantified signal intensities are linearly related to the expression of the corresponding genes in the sample. However, there are sources of nonlinearity in microarray expression measurements. Such nonlinearity study in the expressions of the RNA mixtures provides a new way to analyze gene expression data, and we argue that the nonlinearity can reveal novel information for microarray data analysis. Therefore, we proposed a statistical model, called proportion model, which is based on the linear regression analysis. To approximately quantify the nonlinearity in the dilution design, a new calibration, beta ratio (BR) was derived from the proportion model. Furthermore, a new adjusted fold change (adj-FC) was proposed to predict the true FC without nonlinearity, in particular for large FC. RESULTS: We applied our method to one microarray dilution dataset. The experimental results indicated that, to some extent, there are global biases comparing with the linear assumption for the significant genes. Further analysis of those highly expressed genes with significant nonlinearity revealed some promising results, e.g. 'poison' effect was discovered for some genes in RNA mixtures. The adj-FCs of those genes with 'poison' effect, indicate that the nonlinearity can be also caused by the inherent feature of the genes besides signal noise and technical variation. Moreover, when percentage of overlapping genes (POG) was used as a cross-platform consistency measure, adj-FC outperformed simple fold change to show that Affymetrix and Illumina platforms are consistent. AVAILABILITY: The R codes which implements all described methods, and some Supplementary material, are freely available from http://www.utdallas.edu/~ying.liu/BetaRatio.htm