: Molecular biology data are placed in different databases, repositories and flat files, usually distributed over the web. Distinct data models with schemas that are often changing implement these heterogeneous data sources. It is very important to gather information about these data sources, including schemas and ontology. The usual approach to handle this information integration problem is to use a single model that captures all the needed data and related methods. Instead, this work proposes the use of a domain specific framework for molecular biology data access and applications. This way we can capture multiple schemas and preexisting data sources, besides having a tool for schema evolution maintenance and database instantiation.