An evaluation for merging signaling pathways by using protein-protein interaction data

It has been challenging to develop enhanced pathway tools and methods that could expand the coverage and improve the quality of existing annotated human pathway data. We aim to quantitatively evaluate the processes of merging similar or functionally-related signaling pathways together by linking them with protein-protein interaction (PPI) data. We presented a concept of pathway mergeability to examine the merging potential between two different pathways. We analyzed the mergeability variation of existing pathways in an integrated human pathway database (HPD) and potential pathways expended from a human annotated and predicted protein interaction (HAPPI) database with confidence score for each interaction. Furthermore, by comparing the mergeability variation between expanding existing pathways in the HAPPI and expanding existing pathways in randomly-permuted PPI networks, we revealed a quantitative relationship between signaling pathway data and high-quality PPI data. This quantitative relationship will further guide pathway merging processes and also pathway tool development.