PMC:539999 / 11865-17536 JSONTXT

{"project":"2_test","denotations":[{"id":"15608239-14732688-76592408","span":{"begin":4611,"end":4613},"obj":"14732688"},{"id":"15608239-14980218-76592409","span":{"begin":4646,"end":4648},"obj":"14980218"},{"id":"15608239-12566409-76592410","span":{"begin":4673,"end":4675},"obj":"12566409"}],"text":"dbERGE II DESCRIPTION AND STRUCTURE\n\nQuery form and results pages\nThe general aim of the dbERGE II database, available at http://www.bx.psu.edu/, has not changed with this database upgrade, i.e. to store detailed data from various types of experiments such as DNA transfer (transfections and transgenic mice), binding assays (gel shift, in vitro and in vivo footprints and methylation interference), DNase hypersensitivity and chromatin immunoprecipitation microarray (ChIP-chip) experiments. However, almost every other aspect of this database has changed since it was last reported in 1998. For example, the nested tables of the original database are now in a relational database that can be queried with a graphical query interface similar to that of the GALA database, rather than having to write arcane programmed queries to access the full capability of the database. The new guided, step-wise query interface enables both simple and complex querying capacities for users with varying levels of experience without the need to learn a programming language.\nThe query interface has been redesigned for user convenience. Each query is a step-by-step process that guides a user in a simple and efficient way through the large number of query options. The interface provides complex querying capacity while maintaining an intuitive and non-intimidating approach by splitting the query process into stages. Queries can be constructed to narrow-down and refine results from any data-type recorded in the database, including type of assay, tissue type, protein name, expression levels, author, etc. Experimental regions are recorded with respect to DNA ranges or chromosome locations on the genome. Thus, data are easily output to linked databases like GALA and the UCSC Genome Browser, and upgraded between genome assemblies.\nOnce the query choices are selected, a user is prompted to select the output format. Output choices can be text-based, graphical or the results can be uploaded into other databases. For instance, DNA transfer experiments are displayed graphically in the UCSC Genome Browser as DNA segments corresponding to the genomic regions used in each transfer experiment, flanked by the expression level and type of tissue. Multiple constructs are stacked in a custom browser track for visual comparison with genomic annotations such as the position of known genes. Viewing several tracks simultaneously allows a user to see multiple experimental features such as hypersensitive sites, expressed genes and functional promoters, in order to build evidence for the location of functional elements. These views can be scaled to show varying levels of detail through the variety of query display options, such as ‘dense, packed and full’ offered in the Genome Browser (http://genome.cse.ucsc.edu/). Textual displays allow a choice of the amount of detail to be returned in the output file, describing information about the references, DNA transfer experiments and binding assays. Users wishing to parse large data files can soon request XML output as a form of computer readable, structured results. A schematic illustration of the types of experimental data, the data sources and output options is illustrated in Supplementary Figure 2.\nIn addition to detailed experimental results, the database accepts and displays a second type of data known as a summary. This simplified data type is optimal for entering high-throughput data along with general conclusions, such as ‘hypersensitive site’ ‘functional promoter’, etc., without all the experimental details. Additional bits of information can be included, such as tissue type. Summaries can be queried as a group or by limiting the query to those that fall within a specific genomic locus.\n\nData entry\nData is entered into the dbERGE II database through a graphical data entry interface with permission, or through a more automated manner using a formatted spreadsheet. Individuals working on expression assay studies are encouraged to load their data into the dbERGE II database.\n\nHistory page\nThe dbERGE II database has a user history page analogous to the one in GALA. It stores results of simple queries and allows a user to combine them as unions, intersections or exclude them with a ‘NOT’ statement. Formats for output files created by compound queries are the same as for simple queries. Each query listed on the history page is associated with an automatically generated query description and a count of the number of experiments returned from the query.\n\nData\nThe database holds several types of data including DNase hypersensitive sites (21), ChIP–chip binding sites data (22), functional promoters (23), locus-specific experimental data from the β-globin and CFTR loci, with plans to house non-microarray data generated from the Encyclopedia of DNA Elements (ENCODE) project.\n\nFuture prospects\nIn addition to the features described, the database is scalable to include more experiment types and species. Current efforts are aimed at reducing response times, expanding experimental data types and interaction with other databases, such as Array Express (http://www.ebi.ac.uk/arrayexpress/) and GEO (http://www.ncbi.nlm.nih.gov/geo/), which hold microarray data. Additionally, data from the ENCODE project (http://www.genome.gov/ENCODE/) can be handled and displayed by the dbERGE II database. Most data types (http://www.genome.gov/11009066) are already implemented, such as DNase hypersensitivity, ChIP-chip analyses, enhancer and promoter assays, etc. Furthermore, having links to GALA and UCSC for the comparative display of data will help to analyze the biological implications of the results.\n"}