Netpredictor: R and Shiny package to perform drug-target network analysis and prediction of missing links

Area of Science:

• Bioinformatics
• Computational Biology
• Network Science

Background:

• Biological networks, including unipartite and bipartite types, are crucial for understanding molecular interactions.
• Predicting missing links in these networks is essential for uncovering novel biological functions and relationships.

Purpose of the Study:

• To introduce Netpredictor, an R package designed for the prediction of missing links in unipartite and bipartite networks.
• To provide tools for network property computation, community visualization, and statistical analysis of interactions.

Main Methods:

• Utilizes Random Walk with Restart (RWR) and Network Inference algorithms for missing link prediction.
• Incorporates permutation-based testing for significant interaction identification.
• Supports enrichment analysis for diseases, pathways, and ontologies.

Main Results:

• Random Walk with Restart demonstrated superior performance compared to other algorithms in small dataset comparisons.
• The Netpredictor package facilitates network-based prediction for unipartite and bipartite networks.
• Enrichment analysis using predicted links aids in understanding protein functionality within interactomes.

Conclusions:

• Netpredictor offers a robust platform for network analysis and missing link prediction in biological systems.
• The associated R Shiny web application enhances accessibility for non-programmers.
• Future updates aim to refine algorithms for more streamlined data analysis by scientists.