Network-based association analysis to infer new disease-gene relationships using large-scale protein interactions

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Understanding disease-gene associations is crucial for complex disease research.
• Existing experimental methods for identifying disease-gene links are costly and time-consuming.
• Computational approaches are needed to expand the discovery of potential disease-gene associations.

Purpose of the Study:

• To develop a novel network-based algorithm, Disease-Gene Association (DGA), for predicting disease-gene associations.
• To calculate association scores between genes and diseases using protein interaction network data.
• To identify novel and plausible disease-gene pairs for further investigation.

Main Methods:

• Construction of a large-scale protein interaction network.
• Calculation of relationships between interacting proteins concerning disease relevance.
• Application of the DGA algorithm to score and identify novel disease-gene pairs using neighboring protein information.

Main Results:

• The DGA algorithm achieved high performance in disease-gene prediction, with an F-measure of 0.78 and an AUC of 0.86.
• Identified promising disease-gene associations related to inflammation, lipid metabolism, and various cancers.
• Highlighted key genes for blistering skin diseases and muscular dystrophy.

Conclusions:

• The developed DGA algorithm is a simple and efficient tool for identifying disease-gene associations within protein-protein interaction networks.
• This method provides valuable insights into disease mechanisms and can be generalized to other association studies.
• The findings contribute to advancing biomedical science through enhanced understanding of gene-disease relationships.