HALD, a human aging and longevity knowledge graph for precision gerontology and geroscience analyses

Area of Science:

• Gerontology and Geroscience
• Biomedical Informatics
• Computational Biology

Background:

• Human aging increases susceptibility to age-related diseases, necessitating comprehensive understanding of aging mechanisms.
• Biomedical knowledge on aging and longevity is vast and fragmented across millions of publications, hindering research.
• A unified, accessible knowledge base is crucial for advancing precision gerontology and geroscience.

Purpose of the Study:

• To develop HALD, a text mining-based human aging and longevity dataset.
• To integrate scattered biomedical knowledge into a comprehensive knowledge graph.
• To facilitate precision gerontology and geroscience analyses.

Main Methods:

• Utilized text mining and state-of-the-art Natural Language Processing (NLP) techniques.
• Extracted entities (genes, diseases, biomarkers, etc.) and relations from PubMed literature.
• Constructed a biomedical knowledge graph focused on human aging and longevity.

Main Results:

• HALD contains 12,227 entities across 10 types and 115,522 relations.
• The dataset includes 1,855 aging biomarkers and 525 longevity biomarkers.
• Data was curated from 339,918 biomedical articles up to September 2023.

Conclusions:

• HALD provides a valuable, integrated resource for human aging and longevity research.
• The dataset enhances the accuracy and coverage of knowledge for geroscience.
• HALD supports the development of novel anti-aging therapies and precision medicine approaches.