Semantic Annotation of NIH Funding Data for Supporting Rare Disease Research

Area of Science:

• Biomedical Informatics
• Rare Disease Research
• Health Services Research

Background:

• The number of rare disease research projects has significantly increased in the last two decades.
• Systematic analysis of NIH-funded projects is crucial for assessing research status and identifying gaps.
• Previous work established a knowledge graph for NIH-funded rare disease research based on project titles.

Purpose of the Study:

• To expand the utility of NIH funding data for rare disease research.
• To identify rare disease-related projects using a novel NLP package.
• To semantically annotate project titles and abstracts with biomedical concepts to clarify research aims.

Main Methods:

• Application of the NormMap NLP package to identify rare disease-focused NIH-funded projects.
• Semantic annotation of project titles and abstracts using the Unified Medical Language System (UMLS) biomedical concepts.
• Leveraging extracted information to inform the development of an updated knowledge graph.

Main Results:

• Successfully identified and semantically annotated NIH-funded rare disease research projects.
• Enhanced data representation through NLP and UMLS concept mapping.
• Laid the groundwork for an updated knowledge graph with richer semantic information.

Conclusions:

• The NLP-driven semantic annotation approach effectively enriches NIH funding data.
• This enhanced data facilitates a deeper understanding of the rare disease research landscape.
• The updated knowledge graph is poised to advance rare disease research by highlighting key areas and potential gaps.