CoNI

CoNI integrates numerical omics data in R using partial correlations to detect confounding relationships among paired dependent variables and to construct network representations for molecular network analysis.

Key Features:

• Implementation: Implemented in R for computational analysis of numerical omics datasets.
• Unsupervised integration: Uses partial correlations to identify potential confounding variables among paired dependent variables and to combine two distinct omics datasets into a unified analysis.
• Network representations: Constructs weighted undirected graphs, bipartite graphs, and hypergraph-like structures to represent integrated relationships.
• Hypergraph-like modeling: Integrates datasets into hypergraph-like structures to capture multi-way and higher-order interactions within molecular data.
• Candidate prioritization: Supports identification and ranking of priority biological candidates within integrated networks.
• Comparative network analysis: Enables comparison of network structures across different experimental conditions.

Scientific Applications:

• Molecular Interaction Networks: Representation and analysis of molecular interaction networks using multiple graph models.
• Biological Insight Extraction: Prioritization of candidate molecules or features that may be biologically important within integrated omics data.
• Comparative Analysis: Assessment of how network structures and interactions change under different experimental conditions.

Methodology:

Integration of two numerical omics datasets via partial correlations to detect confounding among paired dependent variables and construction of integrated networks represented as weighted undirected graphs, bipartite graphs, or hypergraph-like structures.