A key goal of bioinformatics is to create database systems and software platforms capable of storing and analysing large sets of biological data. Hundreds of biological databases are now available and provide access to huge amount of biological data. SGD, Yeastract, CYGD-MIPS, BioGrid and PhosphoGrid are five of the most visited databases by the yeast community. These sources provide complementary data on biological entities. Biologists are brought systematically to query these data sources in order to analyse the results of their experiments. Because of the heterogeneity of these sources, querying them separately and then manually combining the returned result is a complex and laborious task. To provide transparent and simultaneous access to these sources, we have developed a mediator-based system called YeastMed. In this paper, we present YeastMed focusing on its architecture.