Multi-Modal Protein Knowledge Graph Construction and Applications (Student Abstract)

Existing data-centric methods for protein science generally cannot sufficiently capture and leverage biology knowledge, which may be crucial many tasks. To facilitate research in this field, we create ProteinKG65, a knowledge graph science. Using gene ontology Uniprot base as basis, transform integrate various kinds of with aligned descriptions sequences, respectively, to GO terms entities. ProteinKG65 is mainly dedicated providing specialized graph, bringing the Gene Ontology function structure prediction. We also illustrate potential applications prototype. Our dataset can downloaded at https://w3id.org/proteinkg65.