CONSTRUCTION OF A NEW DATA MODEL FOR STORING AND RETRIEVING TOXIN INFORMATION

FARIA-CAMPOS A. C. (1), RATES B.(2,3), GOMES R. R. (1), FERNANDES-RAUSCH H. (4), MORATELLI F. S. (1), DE LIMA M. E. (2,3), PIMENTA A. M. C. (2,3), CAMPOS S. V. A. (4), FRANCO, G. R. (1).

(1) Lab. de Genética Bioquímica, (2) Lab. de Venenos e Toxinas Animais, (3) Núcleo de Estudo e Função de Biomoléculas, Dep. de Bioquímica e Imunologia, ICB, (4) Lab. de Universalização de Acesso a Internet, Dep. de Ciência da Computação, ICEx,UFMG.  

Experiments regarding the analysis of toxins from the poisonous arthropod Scolopendra viridicornis (Centipede, Scolopendromorpha) have been performed by members of our group resulting in the production of an expressive amount of data. The analysis of this data has been made manually and the data stored in flat-files or formats associated to a given HPLC or MS equipment. This implies in an increased time for data retrieval and limits data integration to the relationships assigned manually by the researcher. Furthermore, there is no automatic integration between experimentation, analysis and previously existing data. In this work we propose the construction of a data model to store proteomic data using a relational database. The use of a relational database will allow us to store raw and processed data and to represent their relationship. By using this we intend to enable a semi-automatic analysis, recording results in the database and notifying researchers when certain criteria have been met, such as the identification of compounds with the same molecular mass in different experiments. The database construction using a relational approach speeds up the whole process of toxin identification, since it makes explicit associations provided by the researcher at different moments or from experiments of different types. The proposed model uses groups of tables for each data subtype, which store raw data, details regarding the experimental procedure, analysis results and linked publications. The model has been implemented using MySQL under a Linux platform and harbors data from S. viridicornis and other poisonous arthropods. Using this data model we hope to contribute for the study of animal toxins by making the analysis of toxin data easier, faster and more precise.

KEY WORDS: Scolopendra, toxin, database, proteomics.

FINANCIAL SUPPORT: CAPES, CNPq, FAPEMIG.  

CORRESPONDENCE TO: Alessandra Faria-Campos, Email: alessa@icb.ufmg.br