A semantic relationship mining method among disorders, genes, and drugs from different biomedical datasets

Area of Science:

• Biomedical Informatics
• Semantic Web Technologies
• Data Mining

Background:

• Semantic web technologies are widely used in biomedical informatics, with many datasets available in Resource Description Framework (RDF) format.
• Mining semantic relationships among genes, disorders, and drugs is crucial for precision medicine and drug repositioning.
• Existing studies often focus on single datasets, hindering the discovery of current relationships distributed across heterogeneous sources.

Purpose of the Study:

• To develop and present a novel algorithm for mining semantic relationships among genes, disorders, and drugs from multisource heterogeneous biomedical datasets.
• To address the challenge of integrating and querying distributed biomedical data for relationship discovery.

Main Methods:

• Biomedical datasets were converted into RDF triple data and integrated into a storage system using a data integration algorithm.
• Nine query patterns were designed to explore relationships among genes, disorders, and drugs across different datasets.
• A gene-disorder-drug semantic relationship mining algorithm was developed to query these relationships.

Main Results:

• The study successfully mined 25 new gene-disorder-drug relationships, focusing on Parkinson's disease.
• The proposed method demonstrated significant advantages in mining and integrating multisource heterogeneous biomedical datasets.
• Precision increased by 2.51%, query results by 7.7%, and correct queries by 9.5% compared to a previous method.

Conclusions:

• The developed algorithm effectively mines and integrates multisource heterogeneous biomedical datasets.
• The findings facilitate the discovery of current disorder-gene-drug relationships, advancing precision medicine and drug repositioning.
• The method shows improved performance in terms of precision and query accuracy over existing approaches.