Network Medicine to Identify Disease Mechanisms & Treatments

The goal of this project is to use data integration and network analyses to discover disease mechanisms and potential treatments.

  • Wenzhong Xiao, PhD
  • Jingcheng Yang, PhD
  • Li-Yuan Hung / Chanshuo Wu, PhD
  • Martha Eckey, Pharm D
  • Peng Li, PhD
  • Gonghua Li, PhD
  • Feifei Han, PhD
Study Hypothesis and Description

Disease knowledge-based networks reflect the current understanding of diseases between clinical phenotypes/symptoms, drugs, genotypes and other molecular interactions.

By applying network of biomedical knowledge to integrate clinical phenotypes and molecular signatures of ME/CFS, we hope to uncover disease gene modules and to prioritize drug candidates for repurposing to help disease symptoms and progression.

OBJECTIVES

Two researchers in a lab near equipment. Each looks at data on a screen.

  1. Develop a knowledge base of curated studies and the gene-disease-symptom map of ME/CFS available to the research community through the established ME/CFS Data Center.
  2. Conduct a deep-learning heterogeneous networks method for network medicine (MINDR) to discover disease modules and identify drug targets.
  3. Through the computational analysis, identify molecular modules underlying ME/CFS and prioritize in silico drug molecules for repurposing in ME/CFS
  4. Validate the top candidates by evaluation of their reported effects in real world patient data and recommend at least two drugs as candidates for pilot clinical trials.
  5.  
  • We identified diseases similar to ME/CFS and key genes involved.
  • We identified target genes for modulation and are investigating evidence in preclinical animal models and real-world evidence from electronic health records and patient self-reported outcomes.
  • We completed a patient treatment survey of ~4,000 ME/CFS and Long COVID patients on their self-reported outcomes of 150+ treatments. A manuscript has been published: Eckey, M., et al. (2025). Patient-reported treatment outcomes in ME/CFS and long COVID. PNAS. 122(28): e2426874122. doi:10.1073/pnas.2426874122
  • We identified candidate treatments and are discussing with potential partners and collaborators on follow-up human trials and preclinical studies.
  • We continue to retrain the model as new data emerges and identify changes to the results of the model based on the new information.