1. Interaction Network Analysis Using Semantic Similarity Based on Translation Embeddings
- Author
-
Manzoor Bajwa, Awais, Collarana, Diego, Vidal, Maria-Esther, Acosta, Maribel, Cudré-Maurox, Philippe, Maleshkova, Maria, Pellegrini, Tassilo, Sack, Harald, and Sure-Vetter, York
- Subjects
Interaction networks ,Computer science ,0206 medical engineering ,02 engineering and technology ,Learning algorithms ,Supporting knowledge ,Machine learning ,computer.software_genre ,Semantics ,Drug interactions ,Domain (software engineering) ,Dewey Decimal Classification::000 | Allgemeines, Wissenschaft::000 | Informatik, Wissen, Systeme::004 | Informatik ,Knowledge extraction ,Semantic similarity ,Interaction network ,020204 information systems ,Similarity functions ,Similarity (psychology) ,Knowledge graphs ,0202 electrical engineering, electronic engineering, information engineering ,Knowledge Graphs ,Representation (mathematics) ,Drug-target interactions ,Konferenzschrift ,Semantic Web ,Biomedical domain ,business.industry ,Building blockes ,Association reactions ,Embeddings ,Similarity function ,Similarity Function ,Artificial intelligence ,ddc:004 ,business ,computer ,DrugBank ,embeddings ,020602 bioinformatics - Abstract
Biomedical knowledge graphs such as STITCH, SIDER, and Drugbank provide the basis for the discovery of associations between biomedical entities, e.g., interactions between drugs and targets. Link prediction is a paramount task and represents a building block for supporting knowledge discovery. Although several approaches have been proposed for effectively predicting links, the role of semantics has not been studied in depth. In this work, we tackle the problem of discovering interactions between drugs and targets, and propose SimTransE, a machine learning-based approach that solves this problem effectively. SimTransE relies on translating embeddings to model drug-target interactions and values of similarity across them. Grounded on the vectorial representation of drug-target interactions, SimTransE is able to discover novel drug-target interactions. We empirically study SimTransE using state-of-the-art benchmarks and approaches. Experimental results suggest that SimTransE is competitive with the state of the art, representing, thus, an effective alternative for knowledge discovery in the biomedical domain.
- Published
- 2019