Work is underway to develop a research database that will assist investigators in designing protocols for flow-cytometry immunologic experiments. Flexibility, consistency, cooperation and transparency have been identified as highly desirable goals. The relational model, used for implementing the database, has however some limitations. We are exploiting a structural model of the database to develop application-independent capabilities that address those relational weaknesses. Examples of applying such semantic information in the field of integrity maintenance and data abstraction through generation of objects are presented.