Improving neural protein-protein interaction extraction with knowledge selection. (December 2019)

• Record Type: Journal Article
• Authors: Zhou, Huiwei; Li, Xuefei; Yao, Weihong; Liu, Zhuang; Ning, Shixian; Lang, Chengkun; Du, Lei
• Is Part Of: Computational biology and chemistry. Volume 83(2019)
• Publication Date: 2019-12
• Subjects: PPI extraction -- Knowledge selection -- Mutual attention -- Prior knowledge; Chemistry -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; Biochemistry -- Data processing; Biology -- Data processing; Molecular biology -- Data processing; Periodicals; Electronic journals; 542.85
• DOI: 10.1016/j.compbiolchem.2019.107146
• Languages: English
• ISSNs: 1476-9271

Abstract

Graphical abstract: Highlights: A novel knowledge selection model is proposed for PPI extraction. Our method could selectively combine the prior knowledge with textual information. The proposed approach could further establish the connection between the two proteins. Our method leads to a new state-of-the-art performance. Abstract: Protein-protein interaction (PPI) extraction from published scientific literature provides additional support for precision medicine efforts. Meanwhile, knowledge bases (KBs) contain huge amounts of structured information of protein entities and their relations, which can be encoded in entity and relation embeddings to help PPI extraction. However, the prior knowledge of protein-protein pairs must be selectively used so that it is suitable for different contexts. This paper proposes a Knowledge Selection Model (KSM) to fuse the selected prior knowledge and context information for PPI extraction. Firstly, two Transformers encode the context sequence of a protein pair according to each protein embedding, respectively. Then, the two outputs are fed to a mutual attention to capture the important context features towards the protein pair. Next, the context features are used to distill the relation embedding by a knowledge selector. Finally, the selected relation embedding and the context features are concatenated for PPI extraction. Experiments on the BioCreative VI PPI dataset show that KSM achieves a new state-of-the-art performance (38.08 % F 1-score) by adding knowledge selection.