The knowledge database is thought as a modular structure composed by the following parts: a) A written corpus: made of by multilingual specialized texts on human genome, tagged according to the SGML standard, preprocessed, lemmatized, morphosyntactically parsed, and disambiguated. b) A documentary and factographic database: composed, on the one hand, by the bibliografic data made of from the written corpus, references from the terminology database and other references selected by the subject-field experts; on the other hand, by data from people, institutions, enterprises, products, and methods involved within the subject field. c) Terminology database. d) Ontology: using Ontoterm© management system; concepts were entered from the written corpus and the terminology database.