This paper describes an information technology infrastructure aimed at supporting translational bioinformatics studies that require joint management of phenotypic and genotypic data. In particular, we integrated an electronic medical record with an open-source environment for data mining to create a flexible and easy to use query system aimed at supporting the discovery of the most frequent complex traits. We propose a logical formalization to define the phenotypes of interest; this is translated into a graphical interface that allows the user to combine different conditions relative to the electronic medical record data (e.g., the presence of a particular pathology). The phenotypes are then stored in a multidimensional database. Then, the data mining system engine reads the filtered data from the database and executes dynamic queries for analyzing phenotypic data, presenting the results in a multidimensional format through a simple web interface. The system has been applied in a study on genetically isolated individuals, the Val Borbera project.