A computational framework for detecting inter-tissue gene-expression coordination changes with aging.

Area of Science:

• Genomics
• Computational Biology
• Aging Research

Background:

• Aging is a complex biological process influenced by numerous genes and pathways.
• Existing research often focuses on tissue-specific aging, neglecting inter-tissue molecular pathway interplay.
• Understanding cross-tissue aging mechanisms is crucial for developing effective interventions.

Purpose of the Study:

• To develop and validate a novel computational framework for analyzing age-related gene expression patterns across multiple tissues.
• To identify key inter-tissue genes and biological pathways that characterize aging.
• To explore the coordinated molecular changes underlying aging across different tissue types.

Main Methods:

• An adjusted multi-tissue weighted gene co-expression network analysis was employed.
• Differential network connectivity analysis was performed between different age groups.
• Machine learning models (XGBoost and Random Forest) were utilized with gene expression and pathway scores for classification.

Main Results:

• The Random Forest model achieved high accuracy (AUC < 88%) in predicting age groups, highlighting crucial inter-tissue genes.
• Key aging-related pathways involving lipid metabolism, immune system function, and cell communication were identified across tissues.
• Distinct tissue-specific aging pathways were also detected, underscoring tissue heterogeneity.

Conclusions:

• The developed framework effectively reveals the importance of inter-tissue coordination in the aging process.
• The findings provide valuable insights into aging mechanisms and potential therapeutic targets for promoting healthy aging.
• This approach can guide future research into the systemic nature of aging and the development of targeted interventions.