strenda.org) Initiative (Tipton et al., 2014) which created recommendation for the

publication of enzyme data including minimum information for the description of enzymes and related data. These STRENDA recommendations are already accepted by some biological journals and inserted in the author’s guidelines of these journals. Within the biocuration community which was recently enforced by the LY294002 in vitro foundation of the International Society for Biocuration (http://biocurator.org) there are also initiatives to improve the collaboration between database curators and publishers. The adaption of publications to the needs of the database developers will increase the quality and re-usability of published data. The hope from the database curators’ point of view for future papers would be, for example, the consistent usage of identifiers from standard databases, ontologies and controlled vocabularies for a correct identification of entities of interest. Of course, this would only hold for future publications. The extraction of data from already existing papers will be still a big

challenge, including time-consuming manual curation work. Currently there are no software tools to automatically support the identification of missing or inconsistent data. Another challenge for the extraction of data for a reaction kinetics database like SABIO-RK is the spreading of data through the whole text of the publication. In addition, different formats for the representation of data within the paper (e.g. kinetic parameters Terminal deoxynucleotidyl transferase in tables, figures or text) are difficult NVP-BEZ235 order to handle with automatic extraction methods. To follow up our findings we are planning to start a more comprehensive analysis of publications. In addition, we are considering the labeling of the part of information in the database that was missing from the publication, but

has been investigated and added manually by the curators. We have described the biochemical reaction kinetics database SABIO-RK and the data extraction and curation process used to maintain it. SABIO-RK is a manually curated database containing biochemical reactions and their kinetic properties. The database is established as a data resource for both experimentalists and modellers. Data in SABIO-RK are mainly extracted manually from the literature and stored in a structured and standardized format. The database content comprises the relevant data which are essential to describe the characteristics of biochemical reactions, the corresponding biological source, kinetic properties and experimental conditions. Annotations to controlled vocabularies, ontologies, and external databases allow the comparison and exchange of data. For a high quality data in a database the original source should be comprehensive and complete. Based on our experience, and confirmed by our analysis of a set of SABIO-RK relevant publications, we suggest improvement opportunities for publishing experimental data.