Pathogens and gene product normalization in the biomedical literature

Area of Science:

• Biomedical Informatics
• Computational Biology
• Bioinformatics

Background:

• Accurate identification and normalization of pathogens and their gene products are crucial for biomedical literature curation, especially in infectious disease research.
• Existing methods face challenges in disambiguating various names for bacterial species and their gene products.

Purpose of the Study:

• To develop and evaluate a new approach for normalizing pathogen and gene product names in the biomedical literature.
• To improve the accuracy and efficiency of literature curation for infectious diseases.

Main Methods:

• Utilized an Ontology Look-up Service (OLS) for pathogen name disambiguation.
• Incorporated a Gene Ontology Categorizer (GOCat) into a scoring system for pathogen and genome normalization.
• Employed Gene Normalization (GN) methods.

Main Results:

• Achieved 46% precision for cross-species GN on the silver standard and 27% on the gold standard.
• Demonstrated 95% precision and 93% recall for pathogen normalization.
• GOCat significantly improved pathogen and gene normalization, confirming pathogen identification and prioritizing correct gene identifiers.

Conclusions:

• The proposed approach effectively normalizes pathogen and gene product information in biomedical texts.
• Integration of OLS and GOCat enhances disambiguation and accuracy, particularly benefiting infectious disease literature curation.
• Accurate pathogen identification is key to improving overall normalization effectiveness and resolving gene disambiguation issues.