Ratio adjustment and calibration scheme for gene-wise normalization to enhance microarray inter-study prediction

Unified Transcriptomic Signature of Arbuscular Mycorrhiza Colonization in Roots of Medicago truncatula by Integration of Machine Learning, Promoter Analysis, and Direct Merging Meta-Analysis

Plant root symbiosis with Arbuscular mycorrhizal (AM) fungi improves uptake of water and mineral nutrients, improving plant development under stressful conditions. Unraveling the unified transcriptomic signature of a successful colonization provides a better understanding of symbiosis. We developed a framework for finding the transcriptomic signature of Arbuscular mycorrhiza colonization and its regulating transcription factors in roots of Medicago truncatula. Expression profiles of roots in response to AM species were collected from four separate studies and were combined by direct merging meta-analysis. Batch effect, the major concern in expression meta-analysis, was reduced by three normalization steps: Robust Multi-array Average algorithm, Z-standardization, and quartiling normalization. Then, expression profile of 33685 genes in 18 root samples of Medicago as numerical features, as well as study ID and Arbuscular mycorrhiza type as categorical features, were mined by seven models: RELIEF, UNCERTAINTY, GINI INDEX, Chi Squared, RULE, INFO GAIN, and INFO GAIN RATIO. In total, 73 genes selected by machine learning models were up-regulated in response to AM (Z-value difference > 0.5). Feature weighting models also documented that this signature is independent from study (batch) effect. The AM inoculation signature obtained was able to differentiate efficiently between AM inoculated and non-inoculated samples. The AP2 domain class transcription factor, GRAS family transcription factors, and cyclin-dependent kinase were among the highly expressed meta-genes identified in the signature. We found high correspondence between the AM colonization signature obtained in this study and independent RNA-seq experiments on AM colonization, validating the repeatability of the colonization signature. Promoter analysis of upregulated genes in the transcriptomic signature led to the key regulators of AM colonization, including the essential transcription factors for endosymbiosis establishment and development such as NF-YA factors. The approach developed in this study offers three distinct novel features: (I) it improves direct merging meta-analysis by integrating supervised machine learning models and normalization steps to reduce study-specific batch effects; (II) seven attribute weighting models assessed the suitability of each gene for the transcriptomic signature which contributes to robustness of the signature (III) the approach is justifiable, easy to apply, and useful in practice. Our integrative framework of meta-analysis, promoter analysis, and machine learning provides a foundation to reveal the transcriptomic signature and regulatory circuits governing Arbuscular mycorrhizal symbiosis and is transferable to the other biological settings.

The ROS Wheel: Refining ROS Transcriptional Footprints

In the last decade, microarray studies have delivered extensive inventories of transcriptome-wide changes in messenger RNA levels provoked by various types of oxidative stress in Arabidopsis (Arabidopsis thaliana). Previous cross-study comparisons indicated how different types of reactive oxygen species (ROS) and their subcellular accumulation sites are able to reshape the transcriptome in specific manners. However, these analyses often employed simplistic statistical frameworks that are not compatible with large-scale analyses. Here, we reanalyzed a total of 79 Affymetrix ATH1 microarray studies of redox homeostasis perturbation experiments. To create hierarchy in such a high number of transcriptomic data sets, all transcriptional profiles were clustered on the overlap extent of their differentially expressed transcripts. Subsequently, meta-analysis determined a single magnitude of differential expression across studies and identified common transcriptional footprints per cluster. The resulting transcriptional footprints revealed the regulation of various metabolic pathways and gene families. The RESPIRATORY BURST OXIDASE HOMOLOG F-mediated respiratory burst had a major impact and was a converging point among several studies. Conversely, the timing of the oxidative stress response was a determining factor in shaping different transcriptome footprints. Our study emphasizes the need to interpret transcriptomic data sets in a systematic context, where initial, specific stress triggers can converge to common, aspecific transcriptional changes. We believe that these refined transcriptional footprints provide a valuable resource for assessing the involvement of ROS in biological processes in plants.

Meta-analytic framework for sparse K-means to identify disease subtypes in multiple transcriptomic studies

Disease phenotyping by omics data has become a popular approach that potentially can lead to better personalized treatment. Identifying disease subtypes via unsupervised machine learning is the first step towards this goal. In this paper, we extend a sparse K-means method towards a meta-analytic framework to identify novel disease subtypes when expression profiles of multiple cohorts are available. The lasso regularization and meta-analysis identify a unique set of gene features for subtype characterization. An additional pattern matching reward function guarantees consistent subtype signatures across studies. The method was evaluated by simulations and leukemia and breast cancer data sets. The identified disease subtypes from meta-analysis were characterized with improved accuracy and stability compared to single study analysis. The breast cancer model was applied to an independent METABRIC dataset and generated improved survival difference between subtypes. These results provide a basis for diagnosis and development of targeted treatments for disease subgroups.

MetaKTSP: a meta-analytic top scoring pair method for robust cross-study validation of omics prediction analysis

MOTIVATION

Supervised machine learning is widely applied to transcriptomic data to predict disease diagnosis, prognosis or survival. Robust and interpretable classifiers with high accuracy are usually favored for their clinical and translational potential. The top scoring pair (TSP) algorithm is an example that applies a simple rank-based algorithm to identify rank-altered gene pairs for classifier construction. Although many classification methods perform well in cross-validation of single expression profile, the performance usually greatly reduces in cross-study validation (i.e. the prediction model is established in the training study and applied to an independent test study) for all machine learning methods, including TSP. The failure of cross-study validation has largely diminished the potential translational and clinical values of the models. The purpose of this article is to develop a meta-analytic top scoring pair (MetaKTSP) framework that combines multiple transcriptomic studies and generates a robust prediction model applicable to independent test studies.

RESULTS

We proposed two frameworks, by averaging TSP scores or by combining P-values from individual studies, to select the top gene pairs for model construction. We applied the proposed methods in simulated data sets and three large-scale real applications in breast cancer, idiopathic pulmonary fibrosis and pan-cancer methylation. The result showed superior performance of cross-study validation accuracy and biomarker selection for the new meta-analytic framework. In conclusion, combining multiple omics data sets in the public domain increases robustness and accuracy of the classification model that will ultimately improve disease understanding and clinical treatment decisions to benefit patients.

AVAILABILITY AND IMPLEMENTATION

An R package MetaKTSP is available online. (http://tsenglab.biostat.pitt.edu/software.htm).

CONTACT

ctseng@pitt.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Sensitivity Analysis of the MGMT-STP27 Model and Impact of Genetic and Epigenetic Context to Predict the MGMT Methylation Status in Gliomas and Other Tumors

The methylation status of the O(6)-methylguanine-DNA methyltransferase (MGMT) gene is an important predictive biomarker for benefit from alkylating agent therapy in glioblastoma. Our model MGMT-STP27 allows prediction of the methylation status of the MGMT promoter using data from the Illumina's Human Methylation BeadChips (HM-27K and HM-450K) that is publically available for many cancer data sets. Here, we investigate the impact of the context of genetic and epigenetic alterations and tumor type on the classification and report on technical aspects, such as robustness of cutoff definition and preprocessing of the data. The association between gene copy number variation, predicted MGMT methylation, and MGMT expression revealed a gene dosage effect on MGMT expression in lower grade glioma (World Health Organization grade II/III) that in contrast to glioblastoma usually carry two copies of chromosome 10 on which MGMT resides (10q26.3). This implies some MGMT expression, potentially conferring residual repair function blunting the therapeutic effect of alkylating agents. A sensitivity analyses corroborated the performance of the original cutoff for various optimization criteria and for most data preprocessing methods. Finally, we propose an R package mgmtstp27 that allows prediction of the methylation status of the MGMT promoter and calculation of appropriate confidence and/or prediction intervals. Overall, MGMT-STP27 is a robust model for MGMT classification that is independent of tumor type and is adapted for single sample prediction.

Testing the predictive value of peripheral gene expression for nonremission following citalopram treatment for major depression

Major depressive disorder (MDD) in general, and anxious-depression in particular, are characterized by poor rates of remission with first-line treatments, contributing to the chronic illness burden suffered by many patients. Prospective research is needed to identify the biomarkers predicting nonremission prior to treatment initiation. We collected blood samples from a discovery cohort of 34 adult MDD patients with co-occurring anxiety and 33 matched, nondepressed controls at baseline and after 12 weeks (of citalopram plus psychotherapy treatment for the depressed cohort). Samples were processed on gene arrays and group differences in gene expression were investigated. Exploratory analyses suggest that at pretreatment baseline, nonremitting patients differ from controls with gene function and transcription factor analyses potentially related to elevated inflammation and immune activation. In a second phase, we applied an unbiased machine learning prediction model and corrected for model-selection bias. Results show that baseline gene expression predicted nonremission with 79.4% corrected accuracy with a 13-gene model. The same gene-only model predicted nonremission after 8 weeks of citalopram treatment with 76% corrected accuracy in an independent validation cohort of 63 MDD patients treated with citalopram at another institution. Together, these results demonstrate the potential, but also the limitations, of baseline peripheral blood-based gene expression to predict nonremission after citalopram treatment. These results not only support their use in future prediction tools but also suggest that increased accuracy may be obtained with the inclusion of additional predictors (eg, genetics and clinical scales).

Toxicity prediction from toxicogenomic data based on class association rule mining

While the recent advent of new technologies in biology such as DNA microarray and next-generation sequencer has given researchers a large volume of data representing genome-wide biological responses, it is not necessarily easy to derive knowledge that is accurate and understandable at the same time. In this study, we applied the Classification Based on Association (CBA) algorithm, one of the class association rule mining techniques, to the TG-GATEs database, where both toxicogenomic and toxicological data of more than 150 compounds in rat and human are stored. We compared the generated classifiers between CBA and linear discriminant analysis (LDA) and showed that CBA is superior to LDA in terms of both predictive performances (accuracy: 83% for CBA vs. 75% for LDA, sensitivity: 82% for CBA vs. 72% for LDA, specificity: 85% for CBA vs. 75% for LDA) and interpretability.

Covariance adjustment for batch effect in gene expression data

Batch bias has been found in many microarray gene expression studies that involve multiple batches of samples. A serious batch effect can alter not only the distribution of individual genes but also the inter-gene relationships. Even though some efforts have been made to remove such bias, there has been relatively less development on a multivariate approach, mainly because of the analytical difficulty due to the high-dimensional nature of gene expression data. We propose a multivariate batch adjustment method that effectively eliminates inter-gene batch effects. The proposed method utilizes high-dimensional sparse covariance estimation based on a factor model and a hard thresholding. Another important aspect of the proposed method is that if it is known that one of the batches is produced in a superior condition, the other batches can be adjusted so that they resemble the target batch. We study high-dimensional asymptotic properties of the proposed estimator and compare the performance of the proposed method with some popular existing methods with simulated data and gene expression data sets.

Identification of cytokinin-responsive genes using microarray meta-analysis and RNA-Seq in Arabidopsis

Cytokinins are N(6)-substituted adenine derivatives that play diverse roles in plant growth and development. We sought to define a robust set of genes regulated by cytokinin as well as to query the response of genes not represented on microarrays. To this end, we performed a meta-analysis of microarray data from a variety of cytokinin-treated samples and used RNA-seq to examine cytokinin-regulated gene expression in Arabidopsis (Arabidopsis thaliana). Microarray meta-analysis using 13 microarray experiments combined with empirically defined filtering criteria identified a set of 226 genes differentially regulated by cytokinin, a subset of which has previously been validated by other methods. RNA-seq validated about 73% of the up-regulated genes identified by this meta-analysis. In silico promoter analysis indicated an overrepresentation of type-B Arabidopsis response regulator binding elements, consistent with the role of type-B Arabidopsis response regulators as primary mediators of cytokinin-responsive gene expression. RNA-seq analysis identified 73 cytokinin-regulated genes that were not represented on the ATH1 microarray. Representative genes were verified using quantitative reverse transcription-polymerase chain reaction and NanoString analysis. Analysis of the genes identified reveals a substantial effect of cytokinin on genes encoding proteins involved in secondary metabolism, particularly those acting in flavonoid and phenylpropanoid biosynthesis, as well as in the regulation of redox state of the cell, particularly a set of glutaredoxin genes. Novel splicing events were found in members of some gene families that are known to play a role in cytokinin signaling or metabolism. The genes identified in this analysis represent a robust set of cytokinin-responsive genes that are useful in the analysis of cytokinin function in plants.

Comparison and evaluation of pathway-level aggregation methods of gene expression data

Background

Microarray experiments produce expression measurements in genomic scale. A way to derive functional understanding of the data is to focus on functional sets of genes, such as pathways, instead of individual genes. While a common practice for the pathway-level analysis has been functional enrichment analysis such as over-representation analysis and gene set enrichment analysis, an alternative approach has also been explored. In this approach, gene expression data are first aggregated at pathway level to transform the original data into a compact representation in which each row corresponds to a pathway instead of a gene. Thereafter the pathway expression data can be used for differential expression and classification analyses in pathway space, leveraging existing algorithms usually applied to gene expression data. While several studies have proposed the pathway-level aggregation methods, it remains unclear how they compare with one another, since the evaluations were done to a limited extent. Thus this study presents a comprehensive evaluation of six most prominent aggregation methods.

Results

The compared methods include five existing methods--mean of all member genes (Mean all), mean of condition-responsive genes (Mean CORGs), analysis of sample set enrichment scores (ASSESS), principal component analysis (PCA), and partial least squares (PLS)--and a variant of an existing method (Mean top 50%, averaging top half of member genes). Comprehensive and stringent benchmarking was performed by collecting seven pairs of related but independent datasets encompassing various phenotypes. Aggregation was done in the space of KEGG pathways. Performance of the methods was assessed by classification accuracy validated both internally and externally, and by examining the correlative extent of pathway signatures between the dataset pairs. The assessment revealed that (i) the best accuracy and correlation were obtained from ASSESS and Mean top 50%, (ii) Mean all showed the lowest accuracy, and (iii) Mean CORGs and PLS gave rise to the largest extent of discordance in the pathway signature correlation.

Conclusions

The two best performing method (ASSESS and Mean top 50%) are suggested to be preferred. The benchmarking analysis also suggests that there is both room and necessity for developing a novel method for pathway-level aggregation.

Comprehensive literature review and statistical considerations for microarray meta-analysis

With the rapid advances of various high-throughput technologies, generation of ‘-omics’ data is commonplace in almost every biomedical field. Effective data management and analytical approaches are essential to fully decipher the biological knowledge contained in the tremendous amount of experimental data. Meta-analysis, a set of statistical tools for combining multiple studies of a related hypothesis, has become popular in genomic research. Here, we perform a systematic search from PubMed and manual collection to obtain 620 genomic meta-analysis papers, of which 333 microarray meta-analysis papers are summarized as the basis of this paper and the other 249 GWAS meta-analysis papers are discussed in the next companion paper. The review in the present paper focuses on various biological purposes of microarray meta-analysis, databases and software and related statistical procedures. Statistical considerations of such an analysis are further scrutinized and illustrated by a case study. Finally, several open questions are listed and discussed.

A guide to stem cell identification: progress and challenges in system-wide predictive testing with complex biomarkers

We have developed a first generation tool for the unbiased identification and characterization of human pluripotent stem cells, termed PluriTest. This assay utilizes all the information contained on a microarray and abandons the conventional stem cell marker concept. Stem cells are defined by the ability to replenish themselves and to differentiate into more mature cell types. As differentiation potential is a property that cannot be directly proven in the stem cell state, biologists have to rely on correlative measurements in stem cells associated with differentiation potential. Unfortunately, most, if not all, of those markers are only valid within narrow limits of specific experimental systems. Microarray technologies and recently next-generation sequencing have revolutionized how cellular phenotypes can be characterized on a systems-wide level. Here we discuss the challenges PluriTest and similar global assays need to address to fulfill their enormous potential for industrial, diagnostic and therapeutic applications.

Module-based prediction approach for robust inter-study predictions in microarray data

MOTIVATION

Traditional genomic prediction models based on individual genes suffer from low reproducibility across microarray studies due to the lack of robustness to expression measurement noise and gene missingness when they are matched across platforms. It is common that some of the genes in the prediction model established in a training study cannot be matched to another test study because a different platform is applied. The failure of inter-study predictions has severely hindered the clinical applications of microarray. To overcome the drawbacks of traditional gene-based prediction (GBP) models, we propose a module-based prediction (MBP) strategy via unsupervised gene clustering.

RESULTS

K-means clustering is used to group genes sharing similar expression profiles into gene modules, and small modules are merged into their nearest neighbors. Conventional univariate or multivariate feature selection procedure is applied and a representative gene from each selected module is identified to construct the final prediction model. As a result, the prediction model is portable to any test study as long as partial genes in each module exist in the test study. We demonstrate that K-means cluster sizes generally follow a multinomial distribution and the failure probability of inter-study prediction due to missing genes is diminished by merging small clusters into their nearest neighbors. By simulation and applications of real datasets in inter-study predictions, we show that the proposed MBP provides slightly improved accuracy while is considerably more robust than traditional GBP.

AVAILABILITY

http://www.biostat.pitt.edu/bioinfo/

CONTACT

ctseng@pitt.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.