Text Mining for Building Biomedical Networks Using Cancer as a Case Study.

Area of Science:

• Bioinformatics
• Computational Biology
• Text Mining

Background:

• Biological network assembly requires reliable interaction data for accurate system representation.
• High-throughput assays and scientific literature are primary data sources, but literature volume poses curation challenges.
• Manual curation is insufficient for tracking discoveries in the rapidly expanding scientific literature.

Purpose of the Study:

• To review automatic information extraction approaches for enhancing biological networks.
• To highlight the role of text mining and relation extraction in knowledge discovery.
• To focus on state-of-the-art deep learning methods using cancer disease as a case study.

Main Methods:

• Exploration of text mining techniques for knowledge extraction from biomedical documents.
• Focus on relation extraction methods to identify entity interactions.
• Review of deep learning approaches for automatic information extraction.

Main Results:

• Text mining offers an efficient solution for identifying biological interactions from literature.
• Automated tools can lead to more reliable and personalized biological networks.
• Deep learning methods show promise for advancing relation extraction in biomedical text.

Conclusions:

• Automated information extraction is crucial for overcoming the challenges of literature-based biological network construction.
• Advanced text mining, especially deep learning, can significantly improve the accuracy and scope of biological networks.
• This approach is particularly valuable for specific research areas like cancer biology.