Protocol to implement a computational pipeline for biomedical discovery based on a biomedical knowledge graph

Chang Su; Yu Hou; Michael Levin; Rui Zhang; Fei Wang

doi:10.1016/j.xpro.2023.102666

Protocol to implement a computational pipeline for biomedical discovery based on a biomedical knowledge graph

STAR Protoc. 2023 Dec 15;4(4):102666. doi: 10.1016/j.xpro.2023.102666. Epub 2023 Oct 25.

Authors

Chang Su¹, Yu Hou², Michael Levin³, Rui Zhang², Fei Wang⁴

Affiliations

¹ Department of Population Health Sciences, Weill Cornell Medicine, New York, NY 10065, USA. Electronic address: suchangsc10@gmail.com.
² Department of Surgery, University of Minnesota, Minneapolis, MN 55455, USA.
³ Bioengineering Department, College of Engineering, Temple University, Philadelphia, PA 19122, USA.
⁴ Department of Population Health Sciences, Weill Cornell Medicine, New York, NY 10065, USA. Electronic address: few2001@med.cornell.edu.

Abstract

Biomedical knowledge graphs (BKGs) provide a new paradigm for managing abundant biomedical knowledge efficiently. Today's artificial intelligence techniques enable mining BKGs to discover new knowledge. Here, we present a protocol for implementing a computational pipeline for biomedical knowledge discovery (BKD) based on a BKG. We describe steps of the pipeline including data processing, implementing BKD based on knowledge graph embeddings, and prediction result interpretation. We detail how our pipeline can be used for drug repurposing hypothesis generation for Parkinson's disease. For complete details on the use and execution of this protocol, please refer to Su et al.¹.

Keywords: Bioinformatics; Computer Sciences; Health Sciences.

MeSH terms

Artificial Intelligence*
Drug Repositioning
Humans
Parkinson Disease* / drug therapy
Pattern Recognition, Automated

Abstract

MeSH terms

Grants and funding