33 Mining Domain Association Rules from Protein-Protein Interaction Data
-
Published:2006
Download citation file:
Domains are conserved sequence regions in proteins. A protein often contains one to three domains, and the protein function may be inferred from the domain information. While many protein domains have been functionally characterized, the functions of some conserved sequence regions are still poorly understood. The objective of this work is to facilitate the functional annotation of domains using protein-protein interaction data. Our assumption is that if two or more domains co-occur in protein-protein interactions, these domains may be involved in the same biological process. In this work, we first investigate several association rule mining techniques for finding domain correlations. We then propose a new measure, proximity, for mining domain association rules. Finally, we discuss the potential uses of these rules for understanding domain functions.