Analysis of biological processes and diseases using text mining approaches

Area of Science:

• Bioinformatics and Computational Biology
• Genomics and Genetics
• Biomedical Informatics

Background:

• Biomedical text mining complements bioinformatics by extracting biological information from scientific literature.
• Existing literature databases and lexical resources are crucial for effective text mining.
• Information extraction systems are vital for analyzing experimental data and identifying biological relationships.

Purpose of the Study:

• To provide an overview of biomedical text mining systems and their capabilities.
• To describe strategies for linking genes to diseases, including mutations, SNPs, and epigenetic information.
• To demonstrate the utility of text mining in molecular oncology through practical applications.

Main Methods:

• Overview of natural language data characteristics, literature databases, and lexical resources.
• Introduction of selected text mining systems, query types, and result generation.
• Discussion of information extraction for biological relationships (e.g., protein-protein interactions, pathways).
• Description of methods for gene-disease association, mutation, SNP, and methylation mining.
• Implementation of two cancer-related applications: mutation retrieval and breast cancer gene ranking.

Main Results:

• Text mining facilitates the extraction of biologically relevant information, including gene-disease associations and mutations.
• Developed systems can link specific gene mutations to cancer types and rank genes for breast cancer research.
• Community efforts like BioCreative are essential for integrating text mining systems.

Conclusions:

• Biomedical text mining is a valuable tool for molecular oncology and other life science domains.
• Customized text mining systems can be implemented to address specific research questions.
• Future trends highlight the importance of collaborative platforms for advancing text mining capabilities.