Assessment of Modifying versus Non-modifying Protein Interactions

Motivation: The identification of events such as protein-protein interactions (PPIs) from the scientific literature is a complex task. One of the reasons is that there is no formal definition for the syntactical-semantic representation of the relations with which authors of manu-scripts have to comply. In this study, we as-sess the distribution of verbs denoting binary relations between proteins using different cor-pora (AIMed, BioInfer, BioCreAtIve II) for protein-protein interactions and measure their performance for the identification of PPI events (in the BioCreAtIve II corpus) based on syntactical patterns. We distinguish modifying interactions (MIs) such as post-translational modifications (PTMs) from non-modifying in-teractions. We found that MIs are less fre-quent in the corpus but can be extracted at the same precision levels as PPIs. Programmatic access to the text processing modules is available online (www.ebi.ac.uk/ webservices/whatizit/info.jsf