The Pancreatic Expression Database (PED) is a repository for pancreatic-derived -omics data. With a generic web-based system, the database provides the research community with an open access tool to mine currently available pancreatic cancer experimental data sets generated by using large-scale transcriptomic, genomic, proteomic, miRNA and methylomic platforms. Interrogation of the database can be achieved using combined criteria from pancreatic (disease stages, regulation, differential expression, expression, platform technology, publication) and/or public data (antibodies, genomic region, gene-related accessions, ontology, expression patterns, multi-species comparisons, protein data, SNPs). The website also provides users with the opportunity to include their own published dataset in the database.
The first result table provides general information about the gene/miRNA/protein of interest such as HGNC symbol, type (eg., protein coding, non-coding, miRNA), chromosomal position, strand information and corresponding Ensembl transcript/protein id.
The second table contains a summary of the entries associated with the query term stored in the PED. Each entry includes important attributes regarding the study and experiment where the gene/transcript was reported. This includes study type (eg., Transcriptomics, Proteomics, etc.), experimental questions, platform technology, target and baseline specimens/samples used, observations/findings, corresponding fold-change and P-value as well as validation platform.
A link is also provided to download the tables as a simple tab-separated text (TSV) file.
|Summary of Gene Expression/Regulation Events|
|Number of articles||41||18||3||12||3|
|Summary of Copy Number Alteration Events|
|Number of articles||5|
|Copy number altered regions/genes||33092|
|Loss of Heterozygosity||88|
A full list of published datasets and summary of corresponding experimental results are available from PED Mart > Content menu.
The PED database is designed and implemented compliant to the BioMart system. BioMart is a freely available, open source, federated database system that provides unified access to disparate, geographically distributed data sources.
The PED mart allows multiple levels of access. Firstly, access to the PED data is provided through a customized version of MartView, a BioMart web-based query interface based on Perl API. Secondly, the database is available from the BioMart central server where it is exposed to third party software such as the Bioconductor package biomaRt therefore allowing its easy interrogation within the open source R statistical environment and its integration into any expression profiling experiment. Thirdly, the data can be also accessed programmatically through web services. A query constructed in the web-based query interface can be easily converted into an xml or perl template for future bioinformatics expansion and use. Finally, PED is a DAS server providing a Pancreatic Expression DAS annotation available at the Ensembl GeneView.
The query interface to the database is available through navigating to the PED Mart > Database from the menu bar. The interface has two distinct panels on the left and right. The left panel is provided for navigation while user selections take place in the right panel. A summary of user choices is also displayed in the left panel. A simple query involves choosing attributes (or using the default ones) and optional filters if one wants to restrict the query.
First choose a database from the list of databases available. To look specifically at information on pancreatic cancer related study, pick the Pancreatic Expression Database.
Once the database is selected, there are options to pick a relevant dataset from a drop-down list. Here user have options to view information related to Pancreatic gene expression or Pancreatic copy number analysis studies. You can also query other Biomart resources such as Reactome, COSMIC, Interpro and PRIDE.
Once a dataset is selected, the options for choosing filters and attributes are expanded on the left hand side of the page.
Here you can specify the type of study you are interested in and the information you would like to display about a particular study.
A second dataset can be incorporated into the query by clicking on the second Dataset node (if appears) on the left and following the same steps. Linking of the datasets and merging of data into the final result set is handled automatically.
Pancreatic gene expression dataset of PED mart is built on the existing mart structure of Ensembl Genes 63 database. This provides additional flexibility for the users to exploit attributes and filters from the Ensembl Gene system directly into the PED mart.
To choose an attribute simply click on the checkbox next to its description. Sometimes restrictions are imposed in particular section regarding the maximum number of attributes to be selected. If one selects more attributes than permitted, a validation error will occur.
A full list of PED specific attributes and their meanings are available at the Glossary.
There are two different types of filters available in the PED Martview. The first type are Data Filters and generally identified by the Limit to <criteria> filter labels. Data filter allows to extract only matched entries from the database that fits the criteria. The second type are Boolean Filters and generally identified by two radio button options, Only and Excluded, beside the filter option list/text area. The underlying structure of Pancreatic Gene Expression dataset implies that boolean filter allows to extract all entries from the database that associated with the the genes/transcripts that fits the filter criteria.
A full list of PED specific filters and their meanings are available at the Glossary.
To view/save the whole result set, one can choose the appropiate action(s) at the top of the page and click the Go button. However, this can affect how long it takes to retrieve the results. For particularly intensive queries, server time-outs can be a problem, in which case using one of the Web file options is recommended.
After setting-up the appropriate filters from Pancreatic gene expression and/or copy number datasets, users can click on the UCSC button from the toolbar.
A separate browser window will appear where one can view the differentially expressed genes and/or copy number altered regions in the UCSC genome browser under different tracks. User can change the chromosomal view by selecting a chromosome from the drop-down list provided. A simple color-coding scheme is used where up-regulated genes and copy-number gains/amplifications are presented by green, whereas down-regulated genes and copy-number losses/deletions are presented by red. Genes, for which regulation information are not available in PED, are presented by black.
|Track legends (outward→inward)|
|Gene expression tracks|
|Copy number alteration tracks|
When using UCSC genome browser for the visualization purpose, please note that, UCSC custom annotation tracks are viewable on the machine from which PED query is made and kept at least 48 hours after the last access time. Users should be aware that current annotation tracks may contain data from your previous PED queries in the recent history. To view only the annotation data from the current session, go to My Data → Custom Tracks section in the genome browser and delete previous PED annotation tracks.
A summary of the current query in XML format can be obtained by clicking the XML button from the toolbar. This is useful for people accessing BioMart via webservices. Similarly a BioMart perl API script to run the same query can be generated by clicking the Perl button. A permanent url for the current query can be obtained by clicking the URL button. This is particularly useful for bioinformaticians who develop their own tool and wants to to link to PED.
Researchers can now upload their own published datasets through navigating to the PED Mart > Upload from the menu bar. A very basic set of informations are required to complete the process. Whereas most of the elements of the input form are self explanatory, users can get more details about the field by clicking the Information button where available or by consulting this user guide.
There are two different subsets of information that one needs to provide. The first contains information regarding the published article such as title, author, journal etc. One study corresponds to one published article.
It is common that multiple experiments are reported in one article. The second subset contains information regarding an individual experiment such as experiment title, platform technology, specimen details and compiled result file. The submission will not be accepted until the user upload the data file containing the result data.
After the first group of experimental data are submitted, users will be provided options to include additional group of experimental data under the same study, if necessary. User can also start preparing for new study data, if required.
Once the submission is finished, the study information will go through a manual curation/checking phase by our team. The user will be contacted if any issue needs to be resolved. Users will recieve a confirmation email, once their uploaded study data is successfully included in the PED.
Pancreatic expression landscape is a comprehensive study of pancreatic cancer-expression space that integrates data from otherwise disparate sources. The most comprehensive analysis of pancreatic cancer to date, our study primarily serves to highlight limitations inherent with a lack of raw data availability, insufficient clinical/histopathological information and ambiguous data processing. It stresses the importance of a global-systems approach to assess and maximise findings from expression profiling of malignant and non-malignant diseases.
The expression landscape tool can be aby clicking on the Expression Landscape from the menu bar. The page contains brief summary and statistics regarding the studies. The query interface contains four different parts aggregated together which allows users to search using individual or combined criteria as described below.
The first part is a text area where user can provide a list of Affymetrix Genechip Human Genome U133 Plus 2.0 probe id and/or HGNC gene symbols to restrict the query within the provided list. Users can also query by experimental comparisons and set the threshold for logged fold-change and P-value. Users can narrow results to a biological function or canonical pathway of interest.
Based on the search criteria, the matched results will be shown as an HTML table below the query interface. A link will be provided to download the result as a simple tab-separated text (TSV) file.
The table contains general information regarding the matched probes (eg., gene details, genomic mapping) as well as fold-change and P-value observed in the specific expression landscape experiments for that probe. A set of boxplots are provided depicting the probeset expression observed in different types of samples used in the study.
This section provides general information regarding the attributes and filters employed in the PED mart. Similar description are provided for the elements in the Expression Landscape result table.
|Article||Identifier assigned to a particular article.||Anderson NL et al, Freiss H et al, etc.|
|Title||Title of the published article.|
|Source||Name of the journal where the article is published.|
|Volume||Volume and issue number.|
|PMC id||PubMed Central identifier.|
|Technology||Core analysis technology adopted for the experiment.||Transcriptomics|
|Study||Summary that reflects the specific focus of the study.||PDAC study - tissue|
PDAC study - cell line
PDAC study - tissue;Non-invasive detection
PC study- xenograft
|Method||Basic platform technology used in the experiment.||Oligoneucleotide MicroArray hybridization|
miRNA Array hybridization
|Platform||Specific platform technology used in the experiment.||Genechip Human Genome U133A Array (Affymetrix)|
mirVana miRNA Bioarrays (Ambion)
Quantitative methylation-specific PCR (QMSP)
|Comparison||Title of the experiment. It reflects whether the experiment involves differential expression analysis between two groups of samples. It can also be simple expression study within one group of samples.||Pancreatic ductal adenocarcinoma (PDAC) / Normal pancreas (NP)|
Pancreatic Endocrine Tumor (PET) / Normal pancreas (NP)
Adenovirus sensitive / insensitive cell line
Identification of Pancreatic juice proteome
|Validation methods||List of validation platforms used in the study.||Northern blotting|
|GEO/ArrayExpress||GEO or ArrayExpress id.|
|Note (Experiment)||Brief description about the statistical and computational procedure undertaken for the study. It can include the name and verion of the statistical softeware used, quality control and normalization method, threshold used for fold-change or P-value, etc.||VSN normalization;ANOVA;two-sample t-test;P<0.01;|
experiments in triplicate;Quantile normalization;LIMMA;
PRISM;Students t-test;one-way ANOVA;4-fold change;FDR<0.05
DeCyder analysis;two-sided Student's t-test;MASCOT Daemon V2.1;
peptide mass matching (min 5);min 25% sequence identity; etc.
|Specimen/Sample Attributes (Target and Baseline)|
|Type||Primary type of the biological sample.||Bulk tissue|
|Source||Biological source from which the specimen is collected/derived.||PDAC patients|
patients with benign pancreatic disease
PDAC patients with disease-free interval <12 months
|Tissue lesion state||Lesion state of the specimen (applicable for tissue only).||Normal|
|Histology||Sample diagnosis after histopathological examination.||Pancreatic ductal adenocarcinoma (PDAC)|
Normal pancreas (NP) adjacent to cancer
Chronic pancreatitis (CP)
|Collection method||Short description about the specimen collection procedure.||surgical resection;snap-frozen|
fasting,early morning,mid-stream collection
|Preparation method||Short description about the experiment-specific preparation procedure.||Laser capture microdissection (LCM); pooling|
centrifugation;Liquid chromatography (HPLC);pooling
|Cell line||Specific cell lines used in the experiment.||MiaPaCa-2|
|Induced drugs/therapy||Drugs/chemical agents by which specimens are treated/induced.||EGFR inhibitor (Erlotinib); EGFR inhibitor (Cetuximab)|
Hsp 90 Inhibitor (IPI-504)
Serotype 5 human wild-type adenovirus Ad5
Hh pathway antagonist (HhAntag)
|Cellular component||Cell composition of the prepared sample.||Acinar|
|Cellularity||Approximate purity of the desired cells in the sample. Shown as percentage of the total number of cells.|
|Xenograft||Source of xenograft, if the sample is xenografted tissue.||Derived from patient tumour tissue|
|Size||Number of samples in the experiment group.|
|Note||Brief note on the distribution of samples in different subgroups, where samples are distributed across different genders/lesion states/histological diagnosis.||male:7;female:7;metastasis:13|
|Gene Expression Attributes|
|Probe id||Identifier of the probeset/gene/transcript/protein/miRNA as repoted in the study. In some cases, discontinued/obsolete identifiers are replaced by current identifiers.||225220_at|
|Probe type||Identifier type of the reported probe.||HG U133 Plus 2|
|Regulation/Methylation||For a differential expression or methylation study, status of the reported probe. Not always reported in the article from which data are extracted.||Up|
|Fold change||For a differential expression study, the change in expression observed between two experiment groups. Reported in various formats such as plain ratio, logged ratio, etc.|
|P-value||P-value observed for the given statistical test as described in the Note (Experiment) field.|
|Expressed in||General observation regarding where the specific probe is found to be expressed. Unless otherwise stated, the expression site is assumed to be tissue.||PDAC|
|Validation method||Platform used to validate the specific probe expression.||Northern blotting|
|Note||Short note on the definition of fold-change used in the experiment.||T-statistics|
Log of average expression ratio
Averaged ratio of spot intensity
Ratio of iTRAQ reporter ion intensity
|Copy Number Alteration Attributes|
|CNA id||Unique identifier assigned to the copy-number alteration (CNA) entry.||Harada_ampl_1, Suzuki_gain_11, etc.|
|Chromosome||Chromsome id of the CNA region.||1, 2, 3, ...., X, Y.|
|Start position||Chromsome start position of the CNA region.|
|End position||Chromsome end position of the CNA region.|
|Locus||Cytogenetic mapping.||p34.2, q22.1, etc.|
|Subtype||Copy number alteration type.||Gain|
Loss of Heterozygosity (LOH)
|Copy number||Copy number observed for the altered region.|
|Log2ratio||Log ratio as reported in the study.|
|Frequency||Frequency as reported in the study or deducted from the study data.|
|Target gene (reported)||Target gene(s) in the region as reported in the study.|
|Validation method||Platform technology in which reported region is validated.||PCR, FISH, etc.|
|Observed in specimens||Reported specimen id in which the CNA region are observed.|
|Limit to Technology||Limit search for specific experiment Technology.||Data, single-select|
|Limit to Study||Limit search for selected Study type(s).||Data, multiple-select|
|PDAC study - tissue|
PDAC study - cell line
General PC study - xenograft
Therapeutic treatment study
|Select specific Comparison(s) under the study type.||Boolean, multiple-select|
|Limit to Comparisons||Limit search for specific or list of Comparison(s).||Data, multiple-select|
|Specimen/Sample Filters (Target and Baseline)|
|Limit to Type||Limit search for specific sample Type.||Data, single-select|
|Limit to Tissue lesion state||Limit search for specific Tissue lesion state.||Data, single-select|
|Limit to Histology||Limit search for specific sample Histology.||Data, single-select|
|Induced drug/therapy||Select Induced drugs/therapy||Boolean, single-select|
|Cell line||Select specific or list of Cell line(s).||Boolean, multiple-select|
|Cellular component||Select Cellular component of the samples.||Boolean, single-select|
|Xenograft||Select the type of sample Xenograft.||Boolean, single-select|
|Profiling (Transcriptomics/Proteomics/microRNA/Methylomics/Meta-Analysis) Filters|
|Platform||Select specific Platform technology.||Boolean, single-select|
|Fold Change||Search for genes for which Fold change inforamtion is available.||Boolean|
|Up-regulated||Search for genes which are reported Up-regulated in transcriptomics/proteomics/microRNA studies.||Boolean|
|Down-regulated||Search for genes which are reported Down-regulated in transcriptomics/proteomics/microRNA studies.||Boolean|
|Hyper-methylated||Search for genes which are reported Hyper-methylated in methylation studies.||Boolean|
|Hypo-methylated||Search for genes which are reported Hypo-methylated in methylation studies.||Boolean|
|Validated by||Search for genes which are validated by spcific Validation method.||Boolean, single-select|
|Genes Expressed In|
|Tissue||Search for genes which are reported to be expressed in specific or list of tissue types.||Boolean, multiple-select|
|Cell line||Search for genes which are reported to be expressed in pancreatic cancer cell line or HPDE.||Boolean, single-select|
|Other||Search for genes which are reported to be expressed in specific or list of other types of samples.||Boolean, multiple-select|
|Article||Limit search for specific Article.||Data, single-select|
|Copy Number Alteration Filters|
|Experiment: Study||Select specific Experiment.||Boolean, single-select|
|Experiment: Platform Technologyudy||Select specific Platform technology used.||Boolean, single-select|
|Specimen:Limit to Type||Limit search for specific target sample type.||Data, single-select|
|Specimen:Limit to Histology||Limit search for specific target sample histology.||Data, single-select|
|Specimen:Limit to Baseline||Limit search for specific baseline/control sample type.||Data, single-select|
|CNA Information:Type||Search CNA regions which are reported with specific subType.||Boolean, single-select|
|CNA Information:Log2ratio||Search CNA regions for which Log2ratio data are available.||Boolean|
|CNA Information:Frequency||Search CNA regions for which Frequency data are available.||Boolean|
|CNA Information:Validation method||Search CNA regions which are reported to be validation by specific Validation method.||Boolean, single-select|
|Elements in Expression Landscape Results|
|Probe||Affymetrix Genechip Human Genome U133 Plus 2.0 probeset identifier|
|AGI||Mapped AGI identifier|
|Gene Symbol||HGNC gene identifier|
|Gene Name||Description of the gene|
|Comparison||Title of the differential expression study|
|Log-FC||Logged fold change as identified by GC-RMA normalization and LIMMA|
|Adj.P-value||Benjamini and Hochberg (BH) false discovery rate (FDR) adjusted P-value|
|Box plots||Five sets of box plots representing the expression pattern of the probe in different types of samples (PDAC tissue, cancer cell lines, normal tissue adjacent to cancer, xenografted cell lines, other pancreatic cancers) used in the study.|
|Terms related to Histology|
|F-PET||Functioning Pancreatic endocrine tumor|
|IPMN||Intraductal papillary mucinous neoplasms|
|MCN||Mucinous cystic neoplasms|
|NF-PET||Non-functioning Pancreatic endocrine tumor|
|NP-adj||Normal pancreas adjacent to cancer|
|ND||Normal pancreatic duct|
|PACC||Pancreatic acinar cell carcinoma|
|PanIN||Pancreatic intraepithelial neoplasias|
|PDAC||Pancreatic ductal adenocarcinoma|
|PDEC||Poorly-differentiated endocrine carcinoma|
|PET||Pancreatic endocrine tumor|
|WDEC||Well-differentiated endocrine carcinoma|
|WDET||Well-differentiated endocrine tumor|
|Terms related to Experimental Platform|
|2D-DIGE||Two dimensional Difference gel electrophoresis|
|2D-PAGE||Two dimensional Polyacrylamide gel electrophoresis|
|DDD||Digital Differential Display|
|DIGE||Difference gel electrophoresis|
|ELISA||Enzyme-linked immunosorbent assay|
|FISH||Fluorescence in situ hybridization|
|FT-MS||Fourier-transform mass spectrometry|
|HG U133A||Genechip Human Genome U133A Array (Affymetrix)|
|HG U133B||Genechip Human Genome U133B Array (Affymetrix)|
|HG U133 Plus 2||GeneChip Human Genome U133 Plus 2.0 Array (Affymetrix)|
|HG U95 (A, B, C, D, E)||GeneChip Human Genome U95 Set (Affymetrix)|
|HuGeneFL||GeneChip Human Full Length Array HuGeneFL (Affymetrix)|
|HPLC||High-performance liquid chromatography|
|ISH||In situ hybridization|
|IT-MS||Ion-trap mass spectrometry|
|LTQ||Linear trap quadrupole|
|MALDI||Matrix-assisted laser desorption/ionization|
|MSP||Methylation-specific polymerase chain reaction|
|OSU_CCC||Ohio State University Comprehensive Cancer Center|
|PAGE||Polyacrylamide gel electrophoresis|
|PCR||Polymerase chain reaction|
|QMSP||Quantitative (Real-time) methylation-specific polymerase chain reaction|
|qPCR||Quantitative (Real-time) polymerase chain reaction|
|qRT-PCR||Quantitative (Real-time) reverse transcriptase polymerase chain reaction|
|RT-PCR||Reverse transcriptase polymerase chain reaction|
|SAGE||Serial analysis of gene expression|
|SDS-PAGE||Sodium dodecyl sulfate - Polyacrylamide gel electrophoresis|
|TOF-MS||Time-of-flight mass spectrometry|