A Cartographic Tool to Predict Disease Risk-associated Pseudo-Dynamic Networks from Tissue-specific Gene Expression

Chixiang Chen; Biyi Shen; Lijun Zhang; Tonghui Yu; Ming Wang; Rongling Wu

doi:10.21769/BioProtoc.4583

A Cartographic Tool to Predict Disease Risk-associated Pseudo-Dynamic Networks from Tissue-specific Gene Expression

Bio Protoc. 2023 Jan 5;13(1):e4583. doi: 10.21769/BioProtoc.4583.

Authors

Chixiang Chen^{1

2}, Biyi Shen³, Lijun Zhang⁴, Tonghui Yu⁵, Ming Wang⁴, Rongling Wu⁶

Affiliations

¹ Division of Biostatistics and Bioinformatics, University of Maryland School of Medicine, Baltimore, MD, USA.
² Department of Neurosurgery, University of Maryland School of Medicine, Baltimore, MD, USA.
³ Bristol Myers Squibb, Lawrenceville, NJ, USA.
⁴ Department of Population and Quantitative Health Sciences, School of Medicine, Case Western Reserve University, Cleveland, OH, USA.
⁵ School of Mathematics, Hefei University of Technology, Anhui, China.
⁶ Division of Biostatistics and Bioinformatics, College of Medicine, Penn State College of Medicine, Hershey, PA, USA.

Abstract

Understanding how genes are differentially expressed across tissues is key to reveal the etiology of human diseases. Genes are never expressed in isolation, but rather co-expressed in a community; thus, they co-act through intricate but well-orchestrated networks. However, existing approaches cannot coalesce the full properties of gene-gene communication and interactions into networks. In particular, the unavailability of dynamic gene expression data might impair the application of existing network models to unleash the complexity of human diseases. To address this limitation, we developed a statistical pipeline named DRDNetPro to visualize and trace how genes dynamically interact with each other across diverse tissues, to ascertain health risk from static expression data. This protocol contains detailed tutorials designed to learn a series of networks, with the illustration example from the Genotype-Tissue Expression (GTEx) project. The proposed toolbox relies on the method developed in our published paper ( Chen et al., 2022 ), coding all genes into bidirectional, signed, weighted, and feedback looped networks, which will provide profound genomic information enabling medical doctors to design precise medicine. Graphical abstract Flowchart illustrating the use of DRDNetPro. The left panel contains the summarized pipeline of DRDNetPro and the right panel contains one pseudo-illustrative example. See the Equipment and Procedure sections for detailed explanations.

Keywords: Gene regulatory network; Genotype-Tissue expression; Ordinary differential equations; Quasi-dynamic; Risk prediction; Statistical algorithm.

Grants and funding

P30 AG028747/AG/NIA NIH HHS/United States