Microarray-based Expression Data Research Articles

Robust normalization and transformation techniques for constructing gene coexpression networks from RNA-seq data

BackgroundConstructing gene coexpression networks is a powerful approach for analyzing high-throughput gene expression data towards module identification, gene function prediction, and disease-gene prioritization. While optimal workflows for constructing coexpression networks, including good choices for data pre-processing, normalization, and network transformation, have been developed for microarray-based expression data, such well-tested choices do not exist for RNA-seq data. Almost all studies that compare data processing and normalization methods for RNA-seq focus on the end goal of determining differential gene expression.ResultsHere, we present a comprehensive benchmarking and analysis of 36 different workflows, each with a unique set of normalization and network transformation methods, for constructing coexpression networks from RNA-seq datasets. We test these workflows on both large, homogenous datasets and small, heterogeneous datasets from various labs. We analyze the workflows in terms of aggregate performance, individual method choices, and the impact of multiple dataset experimental factors. Our results demonstrate that between-sample normalization has the biggest impact, with counts adjusted by size factors producing networks that most accurately recapitulate known tissue-naive and tissue-aware gene functional relationships.ConclusionsBased on this work, we provide concrete recommendations on robust procedures for building an accurate coexpression network from an RNA-seq dataset. In addition, researchers can examine all the results in great detail at https://krishnanlab.github.io/RNAseq_coexpression to make appropriate choices for coexpression analysis based on the experimental factors of their RNA-seq dataset.

Machine learning-based microarray analyses indicate low-expression genes might collectively influence PAH disease.

Accurately predicting and testing the types of Pulmonary arterial hypertension (PAH) of each patient using cost-effective microarray-based expression data and machine learning algorithms could greatly help either identifying the most targeting medicine or adopting other therapeutic measures that could correct/restore defective genetic signaling at the early stage. Furthermore, the prediction model construction processes can also help identifying highly informative genes controlling PAH, leading to enhanced understanding of the disease etiology and molecular pathways. In this study, we used several different gene filtering methods based on microarray expression data obtained from a high-quality patient PAH dataset. Following that, we proposed a novel feature selection and refinement algorithm in conjunction with well-known machine learning methods to identify a small set of highly informative genes. Results indicated that clusters of small-expression genes could be extremely informative at predicting and differentiating different forms of PAH. Additionally, our proposed novel feature refinement algorithm could lead to significant enhancement in model performance. To summarize, integrated with state-of-the-art machine learning and novel feature refining algorithms, the most accurate models could provide near-perfect classification accuracies using very few (close to ten) low-expression genes.

Identification of candidate miRNA biomarkers for pancreatic ductal adenocarcinoma by weighted gene co-expression network analysis.

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy with a dismal prognosis which is, among others, due to a lack of suitable biomarkers and therapeutic targets. Previously, basic gene expression analysis methods have been used for their identification, but recently new algorithms have been developed allowing more comprehensive data analyses. Among them, weighted gene co-expression network analysis (WGCNA) has already been applied to several cancer types with promising results. We applied WGCNA to miRNA expression data from PDAC patients. Specifically, we processed microarray-based expression data of 2555 miRNAs in serum from 100 PDAC patients and 150 healthy subjects. We identified network modules of co-expressed miRNAs in the healthy subject dataset and verified their preservation in the PDAC dataset. In the non-preserved modules, we selected key miRNAs and carried out functional enrichment analyses of their experimentally known target genes. Finally, we tested their prognostic significance using overall survival analyses. Through WGCNA we identified several miRNAs that discriminate healthy subjects from PDAC patients and that, therefore, may play critical roles in PDAC development. At a functional level, we found that they regulate p53, FoxO and ErbB associated cellular signalling pathways, as well as cell cycle progression and various genes known to be involved in PDAC development. Some miRNAs were also found to serve as novel prognostic biomarkers, whereas others have previously already been proposed as such, thereby validating the WGCNA approach. In addition, we found that these novel data may explain at least some of our previous PDAC gene expression analysis results. We identified several miRNAs critical for PDAC development using WGCNA. These miRNAs may serve as biomarkers for PDAC diagnosis/prognosis and patient stratification, and as putative novel therapeutic targets.

Expression pattern of immunoglobulin superfamily members in the silkworm, Bombyx mori.

Immunoglobulin superfamily (IgSF) proteins are involved in cell adhesion, cell communication and immune functions. In this study, 152 IgSF genes containing at least one immunoglobulin (Ig) domain were predicted in the Bombyx mori silkworm genome. Of these, 145 were distributed on 25 chromosomes with no genes on chromosomes 16, 18 and 26. Multiple sequence alignments and phylogenetic evolution analysis indicated that IgSFs evolved rapidly. Gene ontology (GO) annotation indicated that IgSF members functioned as cellular components and in molecular functions and biological processes. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis suggested that IgSF proteins were involved in signal transduction, signaling molecules and interaction, and cell communication. Microarray-based expression data showed tissue expression for 136 genes in anterior silkgland, middle silkgland, posterior silkgland, testis, ovary, fat body, midgut, integument, hemocyte, malpighian tubule and head. Expression pattern of IgSF genes in the silkworm ovary and midgut was analyzed by RNA-Seq. Expression of 105 genes was detected in the ovary in strain Dazao. Expression in the midgut was detected for 74 genes in strain Lan5 and 75 genes in strain Ou17. Expression of 34 IgSF genes in the midgut relative to the actin A3 gene was significantly different between strains Lan5 and Ou17. Furthermore, 1 IgSF gene was upregulated and 1 IgSF gene was downregulated in strain Lan5, and 4 IgSF genes were upregulated and 2 IgSF genes were downregulated in strain Ou17 after silkworms were challenged with B. mori cypovirus (BmCPV), indicating potential involvement in the response to BmCPV-infection. These results provide an overview of IgSF family members in silkworms, and lay the foundation for further functional studies.

Insights into TREM2 biology by network analysis of human brain gene expression data

Rare variants in TREM2 cause susceptibility to late-onset Alzheimer's disease. Here we use microarray-based expression data generated from 101 neuropathologically normal individuals and covering 10 brain regions, including the hippocampus, to understand TREM2 biology in human brain. Using network analysis, we detect a highly preserved TREM2-containing module in human brain, show that it relates to microglia, and demonstrate that TREM2 is a hub gene in 5 brain regions, including the hippocampus, suggesting that it can drive module function. Using enrichment analysis we show significant overrepresentation of genes implicated in the adaptive and innate immune system. Inspection of genes with the highest connectivity to TREM2 suggests that it plays a key role in mediating changes in the microglial cytoskeleton necessary not only for phagocytosis, but also migration. Most importantly, we show that the TREM2-containing module is significantly enriched for genes genetically implicated in Alzheimer's disease, multiple sclerosis, and motor neuron disease, implying that these diseases share common pathways centered on microglia and that among the genes identified are possible new disease-relevant genes.

Abstract A207: Whole transcriptome profiling in upper aerodigestive tract cancers

Abstract Background: RNA-seq is a powerful technology that allows to obtain sequence and expression information simultaneously. We applied this technology to 4 upper aerodigestive tract cancers. Methods: RNA from four upper aerodigestive tract tumor specimens was extracted and sequencing libraries constructed. Samples were analyzed using an Illumina Genome Analyzer with a paired end module (36 or 54 base read length). Raw data was processed with a proprietary data pipeline. Potential mutations were filtered using both in house and external bioinformatic solutions, involving previously established databases of mutations (COSMIC) and SNPs (dbSNP), and removal of ancestral alleles identified from multiple sequence alignments. These SNPs were then parsed via in house scripts to determine whether the SNPs were present in coding regions, 3′UTR, 5′UTR, or in splice acceptor/donor sites. The coding SNPs were further parsed to determine which SNPs result in non-synonymous changes. RNA-Seq expression data was analyzed using Partek Genomics Suite and compared to microarray based expression data. Pathway analysis was performed using GeneGO Metacore and DAVID. Results: 500Mb to 2GB of data were obtained per sample. Between 700 and 2000 nsSNPs were identified, as well as a large number of alteration in the 3′ and 5′ untranslated regions. Genetic alterations in several commonly mutated genes were identified including TP53, ErbB2, H-RAS, and Cyclin D1. SNPs were enriched in pathways commonly involved in upper aerodigestive tract tumors including cell cycle control, cytoskeleton, and receptor tyrosine kinases. Furthermore we screened for viral sequences identifying one sample that was HPV16 positive. RNA-seq based and microarray based expression data generally correlated, but significant differences were also observed, that were at least partly due to the limited number of RNA-seq reads. Conclusion: Cancer transcriptome sequencing is a promising and cost effective approach for identifying mutations and obtaining expression analysis simultaneously. Transcriptome sequencing holds promise as a readily available platform for assessing potential treatment targets in a specific tumor. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):A207.

Immunoglobulin superfamily is conserved but evolved rapidly and is active in the silkworm,Bombyx mori

Immunoglobulin superfamily (IgSF) proteins are known for their abilities to specifically recognize and adhere to cells. In this paper, we predicted the presence of 133 IgSF proteins in the silkworm (Bombyx mori) genome. Comparison with similar proteins in other model organisms (Caenorhabditis elegans, Drosophila melanogaster, Anopheles gambiae, Apis mellifera and Homo sapiens) indicated that IgSF proteins are conserved but have rapidly evolved from worms to human beings. However, these proteins are well conserved amongst insects. Silkworm microarray-based expression data showed tissue expression of 57 IgSF genes and microbe-induced differential expression of 37 genes. Based on the expression data, we can conclude that the silkworm IgSF is active.

Expansion and Diversification of thePopulusR2R3-MYB Family of Transcription Factors

The R2R3-MYB proteins comprise one of the largest families of transcription factors in plants. R2R3-MYB family members regulate plant-specific processes, such as the elaboration of specialized cell types, including xylem, guard cells, trichomes, and root hairs, and the biosynthesis of specialized branches of metabolism, including phenylpropanoid biosynthesis. As such, R2R3-MYB family members are hypothesized to contribute to the emergence of evolutionary innovations that have arisen in specific plant lineages. As a first step in determining the role played by R2R3-MYB family members in the emergence of lineage-specific innovations in the genus Populus, the entire Populus trichocarpa R2R3-MYB family was characterized. The Populus R2R3-MYB complement is much larger than that found in other angiosperms with fully sequenced genomes. Phylogenetic analyses, together with chromosome placement, showed that the expansion of the Populus R2R3-MYB family was not only attributable to whole genome duplication but also involved selective expansion of specific R2R3-MYB clades. Expansion of the Populus R2R3-MYB family prominently involved members with expression patterns that suggested a role in specific components of Populus life history, including wood formation and reproductive development. An expandable compendium of microarray-based expression data (PopGenExpress) and associated Web-based tools were developed to better enable within- and between-species comparisons of Populus R2R3-MYB gene expression. This resource, which includes intuitive graphic visualization of gene expression data across multiple tissues, organs, and treatments, is freely available to, and expandable by, scientists wishing to better understand the genome biology of Populus, an ecologically dominant and economically important forest tree genus.

RNAdb 2.0--an expanded database of mammalian non-coding RNAs

RNAdb is a comprehensive database of mammalian non-protein-coding RNAs (ncRNAs). There is increasing recognition that ncRNAs play important regulatory roles in multicellular organisms, and there is an expanding rate of discovery of novel ncRNAs as well as an increasing allocation of function. In this update to RNAdb, we provide nucleotide sequences and annotations for tens of thousands of non-housekeeping ncRNAs, including a wide range of mammalian microRNAs, small nucleolar RNAs and larger mRNA-like ncRNAs. Some of these have documented functions and/or expression patterns, but the majority remain of unclear significance, and include PIWI-interacting RNAs, ncRNAs identified from the latest rounds of large-scale cDNA sequencing projects, putative antisense transcripts, as well as ncRNAs predicted on the basis of structural features and alignments. Improvements to the database comprise not only new and updated ncRNA datasets, but also provision of microarray-based expression data and closer interface with more specialized ncRNA resources such as miRBase and snoRNA-LBME-db. To access RNAdb, visit .

Discovering functional transcription-factor combinations in the human cell cycle

With the completion of full genome sequences and advancement in high-throughput technologies, in silico methods have been successfully used to integrate diverse data sources toward unraveling the combinatorial nature of transcriptional regulation. So far, almost all of these studies are restricted to lower eukaryotes such as budding yeast. We describe here a computational search for functional transcription-factor (TF) combinations using phylogenetically conserved sequences and microarray-based expression data. Taking into account both orientational and positional constraints, we investigated the overrepresentation of binding sites in the vicinity of one another and whether these combinations result in more coherent expression profiles. Without any prior biological knowledge, the search led to the discovery of several experimentally established TF associations, as well as some novel ones. In particular, we identified a regulatory module controlling cell cycle-dependent transcription of G2-M genes and expanded its functional generality. We also detected many homotypic combinations, supporting the importance of binding-site density in transcriptional regulation of higher eukaryotes.