Bioinformatics Advance Access originally published online on September 24, 2007
Bioinformatics 2007 23(21):2934-2941; doi:10.1093/bioinformatics/btm430
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Construction of an open-access database that integrates cross-reference information from the transcriptome and proteome of immune cells
1RIKEN Research Center for Allergy and Immunology, 1-7-22 Suehiro-cho, Tsurumi-ku, Yokohama, Kanagawa 230-0045 and 2Department of Human Genome Technology, Kazusa DNA Research Institute, 2-6-7 Kazusa-Kamatari, Kisarazu, Chiba 292-0818, Japan
*To whom correspondence should be addressed.
 |
ABSTRACT |
|---|
 TOP ABSTRACT 1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Motivation: Although a huge amount of mammalian genomic data does become publicly available, there are still hurdles for biologists to overcome before such data can be fully exploited. One of the challenges for gaining biological insight from genomic data has been the inability to cross-reference transcriptomic and proteomic data using a single informational platform. To address this, we constructed an open-access database that enabled us to cross-reference transcriptomic and proteomic data obtained from immune cells.
Results: The database, named RefDIC (Reference genomics Database of Immune Cells), currently contains: (i) quantitative mRNA profiles for human and mouse immune cells/tissues obtained using Affymetrix GeneChip technology; (ii) quantitative protein profiles for mouse immune cells obtained using two-dimensional gel electrophoresis (2-DE) followed by image analysis and mass spectrometry and (iii) various visualization tools to cross-reference the mRNA and protein profiles of immune cells. RefDIC is the first open-access database for immunogenomics and serves as an important information-sharing platform, enabling a focused genomic approach in immunology.
Availability: All raw data and information can be accessed from http://refdic.rcai.riken.jp/. The microarray data is also available at http://cibex.nig.ac.jp/ under CIBEX accession no. CBX19, and http://www.ebi.ac.uk/pride/ under PRIDE accession numbers 2354–2378 and 2414.
Contact: hijikata{at}rcai.riken.jp
Supplementary information: Supplementary data are available at Bioinformatics online.
|
1 INTRODUCTION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
When a complex biological event is to be explored, a data-driven approach is widely accepted as a powerful alternative to a conventional hypothesis-driven one (Smalheiser, 2002). However, the data-driven approach cannot be achieved without high-quality genomic data. In this regard, DNA microarray technology has opened the way to the comprehensive collection of large amounts of data regarding genome-wide gene expression profiles, and is now recognized as a standard tool for the characterization of biological systems at the mRNA level. The description of biological systems with tens of thousands of different mRNA expression levels is highly sensitive to the state of those systems, and has enabled us to discover a number of mRNA biomarkers for various biological events (Abbas et al., 2005; Komor et al., 2005; Su et al., 2004; Zheng et al., 2006). However, the discovery of such biomarkers does not necessarily address the molecular mechanisms underlying the observed biological events. This is mainly because mRNA levels do not always have direct relevance to biological phenomena in general; mRNA is a template of protein synthesis and is not a functional element in itself in most cases. In this respect, protein profiles must have a greater relevance to biological events than mRNA profiles because proteins govern them directly. Thus, genome-wide characterization of the biological system at the protein level is widely considered to be the next goal to achieve, although this in itself is challenging due to technological limitations faced at present (Cutler, 2003). The low availability of quantitative proteomic data is one of bottlenecks when analyzing a biological system from an ‘omics’ viewpoint. To address this problem, we have recently reported an approach for generating quantitative proteomic maps based on two-dimensional gel electrophoresis (2-DE) and have attempted to expand the protein profiling data (Kimura et al., 2006).
As more mRNA and protein profile data has accumulated, the demand for a cross-referencing database has increased significantly. We have therefore developed a web-based platform for sharing specialized immune cell mRNA and protein profile data. For this study, all of the profile data was newly obtained following well-controlled protocols. Affymetrix GeneChip DNA microarray technology was utilized for obtaining mRNA profiles and 2-DE for quantitative protein profiling. In order to integrate the ‘omics’ data obtained for various immune cells, we annotated each sample with controlled vocabularies and implemented a relational database that included the quantitative mRNA and protein profiling data. The Reference genomics Database of Immune Cells (RefDIC) offers a web-based query interface and user-friendly data visualizing facilities, and also allows all of the raw data to be downloaded. RefDIC could serve as a solid reference for the transcriptome and proteome of immune cells, and hence could greatly facilitate the identification of immunologically important genes and proteins that are involved in various immune responses, through cross-referencing quantitative mRNA and protein profiling data.
|
2 SYSTEMS AND METHODS |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
2.1 Samples of immune cells and tissues
In this study, we analyzed a number of samples for mRNA and/or protein profiling including those derived from various tissues and immune cells extracted from human and mouse biopsies, such as T lymphocytes, B lymphocytes, natural killer cells, natural killer T (NKT) cells, dendritic cells, macrophages, mast cells and intestinal epithelial cells, as well as several popular lymphoid or myeloid cell lines. Detailed information regarding these samples and their preparation are available in sample attribute tables, which are accessible through clicking each sample identifier, termed ‘cellID’, found in the data sets section of the website (http://refdic.rcai.riken.jp/dataset.cgi).
2.2 Extracting data from public databases
For the integration of mRNA and protein profiling data for immune cells, we took advantage of the Entrez Gene database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene) at the National Center for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov) because it provides solid references for various lines of genomic information, such as transcript and protein sequences, and Gene Ontology (GO) terms with unique, stable and traceable gene identifiers (Maglott et al., 2007). The sequence data for transcripts and proteins were extracted from RefSeq, GenBank and UniGene databases maintained by NCBI. Additionally, protein sequence data was extracted from the International Protein Index (IPI, http://www.ebi.ac.uk/IPI/IPIhelp.html) served by the European Bioinformatics Institute (EBI, http://www.ebi.ac.uk/), since IPI is designed for complete non-redundant data sets for human, mouse and rat proteomes (Kersey et al., 2004). The protein domain data was extracted from Pfam (http://www.sanger.ac.uk/Software/Pfam/). Because entries in the RefSeq protein database do not have external links to Pfam, we assigned Pfam domains in the respective RefSeq peptide entries using HMMER (Eddy, 1998), with the threshold E-value set to 0.1. All sequence data for probe sets on Affymetrix GeneChip expression arrays was downloaded from the Affymetrix website (http://www.affymetrix.com/index.affx). Human and mouse genome sequence data (Build 36) was also downloaded from the NCBI website. We collected publicly available microarray experimental data from Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/) and ArrayExpress (http://www.ebi.ac.uk/arrayexpress).
2.3 Microarray experiments
Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and/or an RNAeasy kit (Qiagen, Hilden, Germany). The RNA integrity was assessed using a bioanalyzer (Agilent Technologies Inc., Palo Alto, CA, USA). Samples whose RNA integrity number was greater than 7.0 were used for mRNA profiling by microarray analysis. cDNA synthesis, cRNA amplification, biotinylation and fragmentation were performed with a One-Cycle Target Labeling Kit (Affymetrix, Santa Clara, CA, USA). Twenty micrograms of labeled target RNA was hybridized with Mouse Genome 430 2.0 (Mouse430_2) or 430A 2.0 (Mouse430A_2), or Human Genome U133 Plus 2.0 (HG-U133_Plus2) GeneChip expression arrays (Affymetrix) at 45°C for 16 h, as described in the manufacturer's instructions. Washing stages and streptavidin-phycoerythrin staining were conducted using a GeneChip Fluidics Station (Affymetrix). Subsequently, the chips were scanned using a GeneChip Scanner 3000 (Affymetrix). Array data was normalized using either MAS5 (Hubbell et al., 2002) or gcRMA (Irizarry et al., 2003) algorithms. All of the microarray data were deposited to CIBEX with an accession number of CBX19 (http://cibex.nig.ac.jp/index.jsp).
2.4 Gene annotation of Affymetrix probe sets
Each probe set on the Affymetrix arrays consists of 22 oligonucleotides, each 25-bases long. Half of these probes were designed as a ‘perfect match’ (PM) for a specific transcript. To identify which transcript could be probed by each probe set, the 11 PM sequences were subjected to a Basic Local Alignment Search Tool (BLAST) search (Altschul et al., 1990) against RefSeq (39 179 human and 47 930 mouse cDNA sequences), GenBank (128 863 human and 147 850 mouse cDNA sequences) and UniGene [6 988 853 human and 4 277 970 mouse expressed sequence tags (EST) sequences] databases (updated on 1 November 2006). In this study, when the nucleotide sequences for more than 9 of the 11 probes in the set matched that of a given transcript perfectly, we considered that the probe set targeted this particular transcript. If they matched with a complementary sequence of a given transcript or with multiple transcript sequences originating from different gene loci, or if they failed to match any transcript in any of the databases, then it was concluded that the probe set targeted an antisense transcript of a known gene, or was a cross-hybridizing or non-informational probe set, respectively. Consequently, the same Entrez GeneID was given to a probe set targeting a specific transcribed sequence as for one targeting the corresponding transcript. Based on the results of the BLAST search, we classified all of the probe sets into six categories: categories A, B and C included the probe sets targeting specific RefSeq transcripts, specific GenBank transcripts (i.e. present in GenBank but not in RefSeq), and specific transcripts found only in the UniGene database, respectively; category D included those sets targeting antisense transcripts; category E consisted of cross-hybridizing probe sets and category X comprised non-informational probe sets.
2.5 Two-dimensional gel-based proteome experiments
Quantitative protein profiling by 2-DE followed by gel image analyses and mass spectrometry was performed essentially as described (Kimura et al., 2006) with a few modifications (Kimura et al., manuscript in preparation). Briefly, whole-cell protein samples (250 µg) were separated by isoelectric focusing (IEF) in the first dimension and by SDS-PAGE in the second dimension, and subsequently the proteins in the gel were stained with SYPRO Ruby (Molecular Probes, Eugene, OR, USA). The 2-DE gel images were scanned using a ProXPRESS fluorescent imager (PerkinElmer, Inc., Boston, MA, USA) and analyzed using Progenesis Workstation (Nonlinear Dynamics Ltd, Newcastle upon Tyne, UK). The protein levels for each spot on a given gel were normalized by median centering. In order to identify the proteins in each spot on the gels, the mass spectra of digest fragments originating from the proteins excised from each gel were obtained by peptide mass fingerprinting (PMF) and/or MS/MS methods and were searched against the IPI database (Version 3.21; 51 432 mouse protein sequences) using the MASCOT Version 1.8.06 (Matrix Science, London, UK). To define a spot on a gel that corresponded to one on another gel, all gel images were superimposed. When the position of a spot on a gel matched that of one on another, we considered them provisionally as identical protein spots. Each protein was assigned to one of five classes (A–E) based on reliability of identification, primarily according to the degree of confidence in the database search results for PMF and/or MS/MS analysis data (Kimura et al., 2006). The reliability for classes A (conclusive) and B (most likely) was based solely on the MS data. Protein identification in class C was supported by comigration of the previously identified proteins on 2-DE and was consistent with the predicted MS data for the identified proteins, although the MS data alone did not allow us to identify the protein with sufficient confidence. On the other hand, proteins identified in class D were supported only by their electrophoretic behaviors on 2-DE, while proteins in class E remained uncharacterized even after we combined 2-DE and MS data. All of the proteomic data were deposited into the PRIDE database with accession numbers of 2354–2378 and 2414 (http://www.ebi.ac.uk/pride/).
2.6 Correlation analysis of mRNA and protein profiles
For correlation analysis of mRNA and protein levels from the same gene in different samples, Pearson's correlation coefficients were used. Values for the signal intensity of the probe sets were taken as expression levels of transcripts from the corresponding genes. When gene products from a single gene produced multiple protein spots on a 2-DE gel, the sum of the volumes of these protein spots were taken as the volume of these gene products in this study. When a protein spot was found to include two or more proteins, it was excluded from the analysis. The significance of the correlation coefficients was tested using the one-tailed t-distribution with (n–2) degrees of freedom.
2.7 Web database platform
A web server for the RefDIC is on a machine running CentOS (http://www.centos.org/) and the Apache web server (http://apache.org/). All scripts for data querying, retrieving and visualization were written in Perl. A MySQL 4.1x server (http://www.mysql.com/) is used as the storage engine for the database.
|
3 IMPLEMENTATION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
3.1 Database structure and design
RefDIC segregates data into three components of the MySQL tables: ‘Sample’, ‘Microarray experiments’ and ‘Proteome experiments’ (Supplementary Fig. S1). Each immune cell sample is associated internally with a unique identifier, named cell ID, for the determination of connecting lines of information regarding characteristic features of a given sample and the corresponding quantitative mRNA and protein profiling data. The tables of ‘SampleSource’, ‘SampleTreatment’ and ‘TreatmentDescription’ in the ‘Sample’ component provide records of information regarding the characteristic features of a given sample, using controlled vocabulary to ensure that samples can be distinguished. The ‘Microarray experiments’ component has three tables: ‘MicroarrayExperiment’, ‘MAS5_data’ and ‘gcRMA_data’. The ‘MicroarrayExperiment’ table is a record of information regarding the microarray experimental procedures. The ‘MAS5_data’ and ‘gcRMA_data’ tables are records of the values for the signal intensity of the probe sets on Affymetrix GeneChip arrays. The ‘Proteome experiments’ component has four tables: ‘2DEgelExperiment’, ‘2DEgel_data’, ‘protein_identification’ and ‘spot_matching’. The ‘2DEgelExperiment’ table is a record of the number of protein spots that could be quantified on the 2-DE gel. The ‘2DEgel_data’ table is a record of the information regarding respective protein spots, such as the normalized protein level, isoelectric point, molecular weight, etc. The ‘protein_identification’ table is a record of the information regarding identified proteins for the respective protein spots and the ‘spot_matching’ table is a record of the relationship between corresponding protein spots, identified using the procedure described in the Systems and Methods section. The mRNA and protein profiling data for respective genes from the microarray and 2-DE-based proteomic experiments could be linked together with Entrez GeneIDs as a central hub on the MySQL database.
3.2 Using a web-based query interface to access quantitative data
Figure 1 gives an overview of the web-based query interface and retrieval system and how RefDIC responds when a user submits a query. RefDIC provides mRNA and protein profiling data for the same sample. This data is cross-referenced on the relational database and thus the web interface allows users to visualize both the mRNA and protein profiling data. RefDIC also provides the experimental raw data regarding CEL files for the microarray and mass spectra for the proteomic analyses, respectively. The raw data is accessible in ‘Data downloads’ on the RefDIC website.
|
|
4 RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
4.1 Data statistics for RefDIC
RefDIC is comprised of quantitative transcriptomic and proteomic data for immune cells. For transcriptomic data, a total of 125 and 34 microarray data from 60 to 21 different subsets of immune cells and tissues from mouse and human, respectively (Table 1), have been newly obtained and stored in the database. Table 2 summarizes the results for the annotation and classification of Affymetrix GeneChip probe sets. Based on our annotation system we found that 39 406 and 43 835 probe sets on Mouse430_2 and HG-U133_Plus2, respectively, could detect specific transcripts from 20 623 and 19 300 distinct mouse and human genes, respectively. For proteomic data, 23 2-DE gel images from 21 different subsets of immune cells and 2 subsets of epithelial cells were obtained and stored for the mouse (Table 1). We could identify at least 963 protein spots on each gel image, and 435 proteins in total from distinct genes (Table 3 and Supplementary Table S1). We quantified the amount of protein in these spots by gel image analysis. However, because most of our quantitative proteome data were obtained by a single 2-DE experimental run, the quantitative protein data must be interpreted with caution (Fievet et al., 2004; Gustafsson et al., 2004; Mahon et al., 2001). According to our previous study, the average coefficient of variation for the protein spot quantitative values obtained using essentially same protocol as in this study was 26.8% (range 3.6–105%; Kimura et al., 2006). For convenience, the actual gel images used for quantification are accessible by clicking on matchIDs found in the proteome profile section. The current availability of mRNA and protein profile data can be checked in ‘Statistics’ in the data set section of RefDIC (http://refdic.rcai.riken.jp/dataset.cgi).
|
|
|
4.2 Evaluation of the probe set annotation for Affymetrix GeneChip arrays and their relevance to protein profiling data
As some groups have already pointed out, there are problems associated with Affymetrix GeneChip probe set annotation (Dai et al., 2005; Harbig et al., 2005; Zhang, et al., 2005). Because Affymetrix utilized information that was incomplete at the time of GeneChip design, the current GeneChip probe set design could include some discrepancies when compared to the most up to date genomic data. Furthermore, although the probe annotation data provided by Affymetrix stated that multiple probe sets were designed for the same gene, the relationship among these probe sets was unclear in terms of the data provided. To provide information to enable the interpretation of such relationships, we have re-annotated all of the probe sets based on the results of a BLAST search against transcript sequences in RefSeq, GenBank and UniGene databases, and have classified them into six categories (Table 2). This classification was based on extents of curation for each database: RefSeq database is widely accepted as a manually curated collection of sequences representing genomes and thus contains highly reliable transcript information (Pruitt et al., 2007); GenBank database is a comprehensive public repository of sequences, including data from high-throughput cDNA sequencing projects, submitted by the scientific community (Benson et al., 2007); UniGene database is an informational platform for partitioning GenBank sequences, including EST, into a non-redundant set of gene-oriented clusters in silico (Schuler, 1997). Therefore, the probe sets in category A, which corresponded to transcripts in RefSeq, were most likely to probe mature mRNAs specifically. In contrast, we observed that
44% and 51% of probe sets in the categories B and C on Mouse430_2 and HG_U133_Plus2, respectively, were mapped on non-exonic regions (i.e. introns or downstream of the 3'UTR) of protein-coding RefSeq genes by BLAST searches against the genome sequences. Figure 2A shows the distribution of the mean hybridization signal intensities for probe sets across 119 samples in the categories A, B and C on the Mouse430_2 array. As we expected, the mean hybridization signal intensity for category A (5.56 ± 2.87) was significantly higher than that for B (4.09 ± 2.13) and C (3.6 ± 1.67). We observed similar trends with the data for the HG_U133_Plus2 array (data not shown). These results are consistent with those previously reported (Zhang et al., 2005). Based on our annotation, 10 984 and 12 008 genes were targets for multiple probe sets on Mouse430_2 and HG_U133_Plus2 arrays, respectively. We calculated Pearson's correlation coefficients for the expression patterns in different samples among the probe sets to which we assigned the same gene in a pair-wise fashion, and confirmed that there was a better correlation for pairs of probe sets from category A compared to pairs from the other categories (Fig. 2B).
|
We emphasize that our annotation of the probe sets provides the information, whether or not mature mRNA could be probed, for comparison of mRNA and protein profiles. As expected, we found that the correlation between mRNA and protein levels for 166 genes in which products could be identified in all 2-DE gel samples were improved when the probe sets in the A category were used for this calculation: the mean correlation coefficient was 0.47 ± 0.25, whereas that for categories B or C was 0.26 ± 0.31 (Fig. 2C and Supplementary Table S2). Nevertheless, we found that 61% (98 out of 159) genes have significant correlation (P < 0.01), when we chose category A probe sets for the calculation. The rate of occurrence of genes having significant correlation in this study was slightly higher than previously reported (Kwong et al., 2005; Rautajoki et al., 2004).
4.3 Visualization of the mRNA and protein profiling data on RefDIC
The RefDIC website and database were designed to facilitate exploration of mRNA and protein levels in immune cells on demand. This provides a query interface to enable the visualization of mRNA and protein profiling data in the following three ways:
- RefDIC enables the visualization of mRNA profiling data for multiple genes. One can submit a query with various types of inputs, such as gene symbol, gene/protein names and probe set ID, in order to retrieve the mRNA profiling data from the database (Fig. 3A). When a user submits a query, the RefDIC server returns a list of genes and probe sets matching the query (Fig. 3B). The list has annotation information including the probe set category, and users can select a set of genes, cell types and data processing type. The server then returns a heat map of the mRNA profiling data that the user has selected (Fig. 4A). The profile data can be downloaded as tab-delimited text files. This viewer provides links to the chromosome map where the probe sets mapped (inset, Fig. 4A), and it provides lines of information for interpretation when the mRNA levels are discordant among the probe sets to which the same gene was assigned. The viewer also provides several links to a description of the source of samples or genes.
- RefDIC enables the visualization of protein profiling data for multiple genes, as in the case of the mRNA profiling data. For the protein profile viewer it also provides links to lines of information regarding the protein spot, including apparent isoelectric point, apparent molecular mass, the mass spectral analysis, the reliability of the protein identification and the enlarged gel image (Supplementary Fig. S2A). The server also provides an interactive viewer for a whole 2-DE gel image of a particular sample accessible from ‘2DE-Gel image viewer’ (http://refdic.rcai.riken.jp/2Dgel_viewer.cgi), and enables the user to view and search the information regarding the protein spots by clicking each protein spots on the gel image (Supplementary Fig. S2B).
- RefDIC enables the visualization of both mRNA and protein profiling data for a single gene, as shown in Figure 4B. In this viewer both mRNA and protein levels are represented by the relative amount of each in different samples, and thus the user can easily find at a glance whether their profiles correlate or not. As an example, Figure 4B shows that the ornithine aminotransferase gene demonstrated good correlation between mRNA and protein levels (r2 = 0.88), whereas peroxiredoxin 3 showed poor correlation (r2 = 0.01).
|
|
4.4 Links to the external microarray data on public repositories
To function as a data-sharing platform for genomics information in immunology, RefDIC has a function to search and retrieve the microarray experimental data in the public repositories GEO and Array Express. At present, data searching is confined to data sets using Affymetrix GeneChip Mouse430_2, Mouse430A_2 and HG_U133_Plus2 arrays simply because they are directly comparable with our data. This function would greatly facilitate data-sharing processes in immunology. In the future, we plan to add a tool to enable the visualization of expression profiles for publicly available microarray data that is relevant to immunology, together with our data on the website.
|
5 CONCLUSION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
We have successfully developed a web-based platform for the integration of quantitative immunological transcriptomic and proteomic data. RefDIC was implemented with user-friendly query interfaces and enabled us to search and visualize both mRNA and protein profiling data for various types of immune cells at a glance. All of the information and data, including experimental raw data, are freely available via the Internet. Because this informational platform for cross-referencing data has an expandable structure, it can easily incorporate public data. RefDIC continues to grow in terms of the quantity of mRNA and protein profiling data available, and it continues to integrate more relevant information for genes and proteins, such as cis-elements including promoter sequences, and data for protein–protein interactions.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The authors thank Dr Tomoko Mori, Sachiko Matsuyama, Masako Mori, Shoji Tane and Yuri Ishizu for their help in performing microarray and 2-DE experiments. The authors also thank Yasuaki Murahashi and Miho Izawa for their help in setting up the web server. The first author is grateful to Advanced Center of Computing and Communication, RIKEN for kindly providing the necessary computing resources.
Conflict of Interest: none declared.
|
FOOTNOTES |
|---|
Associate Editor: Dmitrij Frishman
Received on March 7, 2007; revised on July 31, 2007; accepted on August 17, 2007
|
REFERENCES |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSION5 CONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Abbas AR, et al. Immune response in silico (IRIS): immune-specific genes identified from a compendium of microarray expression data. Genes Immun. (2005) 6:319–331.[CrossRef][Web of Science][Medline]
Altschul SF, et al. Basic local alignment search tool. J. Mol. Biol. (1990) 215:403–410.[CrossRef][Web of Science][Medline]
Benson DA, et al. GenBank. Nucleic Acids Res. (2007) 35:D21–D25.
Cutler P. Protein arrays: the current state-of-the-art. Proteomics (2003) 3:3–18.[CrossRef][Web of Science][Medline]
Dai M, et al. Evolving gene/transcript definitions significantly alter the interpretation of GeneChip data. Nucleic Acids Res. (2005) 33:e175.
Eddy SR. Profile hidden Markov models. Bioinformatics (1998) 14:755–763.
Fievet J, et al. Assessing factors for reliable quantitative proteomics based on two-dimensional gel electrophoresis. Proteomics (2004) 4:1939–1949.[CrossRef][Web of Science][Medline]
Gustafsson JS, et al. Statistical exploration of variation in quantitative two-dimensional gel electrophoresis data. Proteomics (2004) 4:3791–3799.[CrossRef][Web of Science][Medline]
Harbig J, et al. A sequence-based identification of the genes detected by probesets on the Affymetrix U133 plus 2.0 array. Nucleic Acids Res. (2005) 33:e31.
Hubbell E, et al. Robust estimators for expression analysis. Bioinformatics (2002) 18:1582–1592.
Irizarry RA, et al. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. (2003) 31:e15.
Kersey PJ, et al. The international protein index: an integrated database for proteomics experiments. Proteomics (2004) 4:1985–1988.[CrossRef][Web of Science][Medline]
Kimura Y, et al. Construction of quantitative proteome reference maps of mouse spleen and lymph node based on two-dimensional gel electrophoresis. Proteomics (2006) 6:3833–3844.[CrossRef][Web of Science][Medline]
Komor M, et al. Transcriptional profiling of human hematopoiesis during in vitro lineage-specific differentiation. Stem Cells (2005) 23:1154–1169.
Kwong KY, et al. Synchronous global assessment of gene and protein expression in colorectal cancer progression. Genomics (2005) 86:142–158.[CrossRef][Web of Science][Medline]
Maglott D, et al. Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res. (2007) 35:D26–D33.
Mahon P, et al. Quantitative and reproducible two-dimensional gel analysis using Phoretix 2D Full. Electrophoresis (2001) 22:2075–2085.[CrossRef][Web of Science][Medline]
Pruitt KD, et al. NCBI reference sequences (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res. (2007) 35:D5–D12.[CrossRef][Web of Science][Medline]
Rautajoki K, et al. Proteome characterization of human T helper 1 and 2 cells. Proteomics (2004) 4:84–92.[CrossRef][Web of Science][Medline]
Schuler GD. Pieces of the puzzle: expressed sequence tags and the catalog of human genes. J. Mol. Med. (1997) 75:694–698.[CrossRef][Web of Science][Medline]
Smalheiser NR. Informatics and hypothesis-driven research. EMBO rep. (2002) 3:702.[CrossRef][Web of Science][Medline]
Su AI, et al. A gene atlas of the mouse and human protein-encoding transcriptomes. Proc. Natl Acad. Sci. USA (2004) 101:6062–6067.
Zhang J, et al. Detecting false expression signals in high-density oligonucleotide arrays by an in silico approach. Genomics (2005) 85:297–308.[CrossRef][Web of Science][Medline]
Zheng C, et al. Gene expression profiling of CD34+ cells identifies a molecular signature of chronic myeloid leukemia blast crisis. Leukemia (2006) 20:1028–1034.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Nishida, A. Hasegawa, S. Nakae, K. Oboki, H. Saito, S. Yamasaki, and T. Hirano Zinc transporter Znt5/Slc30a5 is required for the mast cell-mediated delayed-type allergic reaction but not the immediate-type reaction J. Exp. Med., June 8, 2009; 206(6): 1351 - 1364. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Keerthikumar, R. Raju, K. Kandasamy, A. Hijikata, S. Ramabadran, L. Balakrishnan, M. Ahmed, S. Rani, L. D. N. Selvan, D. S. Somanathan, et al. RAPID: Resource of Asian Primary Immunodeficiency Diseases Nucleic Acids Res., January 1, 2009; 37(suppl_1): D863 - D867. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Kitamura, M. Ito, T. Yuasa, C. Kikuguchi, A. Hijikata, M. Takayama, Y. Kimura, R. Yokoyama, T. Kaji, and O. Ohara Genome-wide identification and characterization of transcripts translationally regulated by bacterial lipopolysaccharide in macrophage-like J774.1 cells Physiol Genomics, October 8, 2008; 33(1): 121 - 132. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Hara, C. Ishihara, A. Takeuchi, L. Xue, S. W. Morris, J. M. Penninger, H. Yoshida, and T. Saito Cell Type-Specific Regulation of ITAM-Mediated NF-{kappa}B Activation by the Adaptors, CARMA1 and CARD9 J. Immunol., July 15, 2008; 181(2): 918 - 930. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||