PathwayFinder: paving the way towards automatic pathway extraction
APBC '04 Proceedings of the second conference on Asia-Pacific bioinformatics - Volume 29
Hi-index | 0.00 |
Although there have been many research projects to extract protein pathways, most such information still exists only in the scientific literature, usually written in natural languages and defying data mining efforts. We present a novel and robust approach for extracting protein-protein interactions from the literature. Our method uses a dynamic programming algorithm to compute distinguishing patterns by aligning relevant sentences and key verbs that describe protein interactions. A matching algorithm is designed to extract the interactions between proteins. Equipped only with a protein name dictionary, our system achieves a recall rate of about 80.0% and a precision rate of about 80.5%.