| Description: |
Introduction
‘Manually Curated Database of Rice Proteins’ is a literature based manually curated protein centric database of rice proteins. The database provides experimental data embedded in published articles in a computer searchable format. The literature curation workflow is fundamentally different than most other literature mining approaches, which concentrate on mining the text of the articles to extract information. The in-house developed manual curation models enable digitization of the experimental data itself. Emphasis is given to experiments that provide direct information about a protein/gene expression or activity. Thus, data from experiments such as quantitative/semi-quantitative RT-PCR, Northern analysis, protein-protein or DNA-protein interaction, enzymatic assays, trait analysis etc. have been manually digitized using these in-house developed data curation models. As a result of such curation one is able to search, for example, all RT-PCR based gene expression data for a particular rice protein published in several different publications in a matter of seconds. The data curation models extensively utilize well-known ontologies such as Gene Ontology (GO), Plant Ontology (PO), Trait Ontology (TO), Environmental Ontology (EO) etc. to facilitate seamless integration of experimental data across publications. Since the existing ontologies were not sufficient to represent the immensely diverse data available in published literature a large number of new terms have also been appended to the existing ontologies. Moreover, several other coding systems were also developed to systematically capture aspects of the experimental data that was beyond the scope of the existing ontologies. Current release of the database has data for over 2390 rice proteins from over 550 research articles.
|
