Knowledge Representation of Protein PTMs and Complexes in the Protein Ontology: Application to Multi-Faceted Disease Analysis

Alterations in protein post-translational modification (PTM) and PTM cross-talk are increasingly being appreciated as driving mechanisms of human disease. The Protein Ontology (PRO) is a valuable resource to study the relation between PTM and disease because it represents individual proteins and protein complex subunits at the proteoform (e.g., isoform, PTM form, and sequence variant) level, with links to their functional properties. We constructed a multi-relation network that represents knowledge obtained from large scale text-mining for phosphorylation-dependent proteinprotein interactions (PPIs) and their disease associations, built on the PRO framework for representation of PTM forms, complexes, and protein families, as well as their attributes and relationships. We then conducted two case studies that demonstrate the use of PRO in disease analysis. (i) We performed cross-species comparisons of two glioma-associated phosphorylated proteoforms of the human DNMT1 methylase, which revealed that the forms are not strictly conserved in mouse, a frequently used glioma model system. (ii) We used PRO-defined proteoforms of the oncoprotein beta-catenin phosphorylated on various combinations of the N-terminal sites, Ser-33, Ser-37, Thr-41, and Ser-45, to interpret a hierarchical clustering analysis of cancer types based on their pattern of mutations in these sites. The cancers formed two major clusters: one with mutations in Ser-33/Ser-37/Thr-41 and the other with mutations in Thr-41/Ser-45. Proteoform-specific annotation in PRO suggests that stabilization of beta-catenin may play a role in oncogenesis in the first group, whereas alterations in beta-catenin transcriptional or cell adhesion activity may play a more important role in the second group. Together, these scenarios illustrate the general applicability of PRO to disease understanding. Future plans include the integration of PRO with other semantic resources to increase our ability to address these problems with computational reasoning. Keywords—Protein Ontology; phosphorylation; text mining; beta-catenin; cancer