Mining web services for pathway discovery

Computer-based pathway discovery is concerned with two important objectives: pathway identification and verification. Conventional mining and modeling approaches aimed at pathway discovery are often effective at achieving either objective, but not both. To achieve both of these objectives, we propose to model biological entities using Web services. We describe our Web service mining framework used for mining pathways linking these entities. As an application, we model Aspirin, COX and IKK-beta related entities and show how pathways involving these entities are identified and statically verified using this mining approach.