Guided assembly of cellular network models from knowledge in literature

Area of Science:

• Computational Biology
• Bioinformatics
• Systems Biology

Background:

• Biological model creation is time-consuming, relying on manual literature review and experiments.
• Existing methods lack efficiency in extracting and integrating knowledge from scientific literature.

Purpose of the Study:

• To develop an automated method for suggesting biological model extensions.
• To leverage published literature for rapid and relevant model updates.

Main Methods:

• Developed a novel automated approach to extract and organize biological events from literature.
• Utilized a collaboration graph with event occurrence frequency as a key metric.
• Assessed the impact of common graph centrality metrics on event assessment.

Main Results:

• The automated method successfully suggests relevant information for model extensions in seconds.
• Achieved an average recall of 82% in identifying desired new events.
• Demonstrated reliability across models for T cell differentiation, T cell large granular lymphocyte, and pancreatic cancer.

Conclusions:

• The proposed automated method significantly accelerates biological model extension.
• This approach offers an efficient way to integrate new knowledge into existing biological models.
• The findings highlight the potential of literature-based computational analysis in systems biology.