A Knowledge Base and Expert System for 2D-PAGE

Jonathan Epstein1, Jonathan Gordin1, Donita Garland2 and Gary Giulian3
1- National Institute of Child Health and Human Development
2- National Eye Institute
3-National Institute of Mental Health



Two-dimensional gel electrophoresis (2D-PAGE) is, despite its pitfalls, still the primary technique for protein identification and separation.  Selecting an appropriate 2D-PAGE protocol requires a level of expertise which is rarely available, despite the ubiquity of the protocol.


We are creating a knowledge base for 2D-PAGE which captures the semantics of sample preparation, the first dimension of separation (by pI) and the second dimension of separation (by molecular weight).  The captured semantics include such factors as voltage application, buffers and other reagents.


Several reported exported protocols have been manually extracted from the scientific and sales literature, and entered into this dataset.  The dataset itself is exportable in DAML+OIL format, which is an XML-based format that is the industry standard for representing complex knowledge bases.


The ultimate goal is to convert this dataset into an expert system, using industry-standard expert system software such as CLIPS and JESS, taking advantage of the pre-existing compatibility between DAML+OIL-friendly software tools such as Protege-2000 and JESS.  One portion of the resulting system will permit a 2D-PAGE domain expert to expand and edit the knowledge base.  The other end of the system will be a web-based interview system which will guide a laboratory scientist as to how to run their gel.