Constructing biological networks through combined literature mining and microarray analysis: a LMMA approach

Area of Science:

• Systems Biology
• Bioinformatics
• Computational Biology

Background:

• Biological network reconstruction is crucial for understanding biological processes and system organization.
• Integrating diverse data sources, such as scientific literature and gene expression data, is essential for comprehensive network analysis.

Purpose of the Study:

• To develop and present a combined Literature Mining and Microarray Analysis (LMMA) approach for constructing gene networks.
• To demonstrate the efficacy of LMMA in creating more reliable biological networks compared to traditional methods.

Main Methods:

• A global gene network is initially constructed using a literature-based co-occurrence method.
• The network is subsequently refined using microarray gene-expression data via a multivariate selection procedure.
• The LMMA approach integrates text mining of literature with quantitative gene expression analysis.

Main Results:

• Application of LMMA to angiogenesis revealed a more reliable gene network than a network derived solely from co-occurrence data.
• The LMMA-based network demonstrated superior performance in handling multiple levels of biological information, including KEGG genes, KEGG Orthology, and pathways.
• The refined network provides a more accurate representation of biological interactions.

Conclusions:

• The combined Literature Mining and Microarray Analysis (LMMA) approach offers a robust method for gene network reconstruction.
• LMMA enhances the reliability and accuracy of biological networks by integrating literature-based and experimental data.
• This approach facilitates a deeper understanding of complex biological systems and processes.