A challenge in computational biology is to uncover thegene-protein inter-actions from gene expression data. A goodapproach for discovering inter-actions between genes based onmultiple expression measurements is to use Bayesian networks. Tothat end, BUBBLE (Bayesian User Based Bio-logical LearningEnvironment) has been developed. BUBBLE allows a re-searcher tocreate pathways, to specify prior constraints, to import data fromvarious databases and learn a Bayesian network model. In addition,User Relevance Feedback queries the user with optimal chosenquestions to improve the correctness of the pathways. The goal ofthe project was to develop an appropriate and intuitive environmentso that the user could be effective at contributing to the Bayesiannetwork construction and analysis.