Pre-training graph neural networks for link prediction in biomedical networks.

Journal: Bioinformatics (Oxford, England)
Published Date:

Abstract

MOTIVATION: Graphs or networks are widely utilized to model the interactions between different entities (e.g. proteins, drugs, etc.) for biomedical applications. Predicting potential interactions/links in biomedical networks is important for understanding the pathological mechanisms of various complex human diseases, as well as screening compound targets for drug discovery. Graph neural networks (GNNs) have been utilized for link prediction in various biomedical networks, which rely on the node features extracted from different data sources, e.g. sequence, structure and network data. However, it is challenging to effectively integrate these data sources and automatically extract features for different link prediction tasks.

Authors

  • Yahui Long
    College of Computer Science and Electronic Engineering, Hunan University, Changsha, 410000, China.
  • Min Wu
    Guizhou University of Traditional Chinese Medicine, Guiyang, Guizhou Province, China.
  • Yong Liu
    Department of Critical care medicine, Shenzhen Hospital, Southern Medical University, Guangdong, Shenzhen, China.
  • Yuan Fang
    Department of Neurology, Dongyang People's Hospital, Affiliated to Wenzhou Medical University, Dongyang, China.
  • Chee Keong Kwoh
    School of Computer Science and Engineering,  Nanyang  Technological  University,  50  Nanyang  Avenue,  639798, Singapore.
  • Jinmiao Chen
    Singapore Immunology Network (SIgN), Agency for Science, Technology and Research, Singapore, Singapore.
  • Jiawei Luo
  • Xiaoli Li
    State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, China.