PMC:540010 / 3000-6382 JSONTXT

{"project":"2_test","denotations":[{"id":"15608260-15033600-77097092","span":{"begin":949,"end":950},"obj":"15033600"},{"id":"15608260-11125075-77097093","span":{"begin":2630,"end":2631},"obj":"11125075"},{"id":"15608260-12613259-77097094","span":{"begin":2649,"end":2651},"obj":"12613259"},{"id":"15608260-12520011-77097095","span":{"begin":3227,"end":3229},"obj":"12520011"},{"id":"15608260-14681407-77097096","span":{"begin":3257,"end":3259},"obj":"14681407"}],"text":"SUBMISSION AND CURATION\nThere are two major submission routes to ArrayExpress: (i) online via the MIAMExpress data submission tool, and (ii) via a MAGE-ML-based pipeline set-up with an external application or database. Currently, more than a half of all submissions have been submitted online. MIAMExpress is primarily aimed at users with no substantial local bioinformatics support and with no access to a local database providing direct deposition. No prior knowledge of the MIAME guidelines is required, as contextual help on the information required and help on the use of MIAMExpress is provided via links from the web interface. Submitters progress through a series of simple web forms to describe their experiment and upload the data files. MIAMExpress is an open source software that can be customized for use by a single laboratory, or for particular application domains. Examples of customization hosted at the EBI include the toxicology (8) and plant-specific MIAMExpress versions. MIAMExpress is now installed in 35 locations worldwide. Source code and installation information can be found at http://sourceforge.net/projects/miamexpress/.\nThe ArrayExpress curation team processes each submission before it is loaded into the repository. Submissions are checked for MIAME compliance, accuracy and completeness of biological information provided, as well as for data consistency (e.g. it is checked if submitted data files match the specified array designs). During the curation process, the curation team may contact the submitter if inconsistencies in the data are found. Once the data are successfully loaded into ArrayExpress repository, the experiment is issued an accession number and a password is provided to the submitter if requested. The data in the repository are owned by the submitter, released on the date specified or upon publication in a journal and no changes are made without the submitter's consent. Where array designs are commercially available these are pre-loaded into ArrayExpress by the curators in response to user request. Information on custom-made arrays is submitted as a tab-delimited file containing position information and annotation information.\nThe ArrayExpress curators work with external databases when setting up a direct data-submission pipeline to ensure that data are MIAME compliant and well formatted. Once a pipeline is established, the submissions are curated at the source database and monitored by ArrayExpress curators. MAGE-ML-based pipelines have been established from 15 external databases, manufacturers or tools, including the Stanford Microarray Database (SMD) (9), MIDAS at TIGR (10), from externally installed MIAMExpress systems at Cambridge University and the European Molecular Biology Laboratory at Heidelberg, as well as the array manufacturers, Affymetrix and Agilent.\nFurther data curation is performed when populating the ArrayExpress warehouse from the repository. The curators select data based on their MIAME compliance, presence of normalized data and the quality of the biological annotation. Array designs are additionally annotated to the current version of the sequence databases at the EBI and up-to-date gene annotation, such as InterPro (11) Gene Ontology (GO) terms (12) and gene names are added, while the original array annotation supporting the publication is maintained in the repository."}