TkNA

TkNA reconstructs causal transkingdom networks from multi-omic datasets to identify and characterize host–microbiota and other inter-omic interactions.

Key Features:

• Holistic Integration: Combines multiple omic layers including genomics, transcriptomics, and proteomics across cohorts for integrated analysis.
• Causal Inference: Applies causality-sensitive metrics and statistical thresholds to emphasize causal relationships over mere associations.
• Network Reconstruction: Builds a statistical transkingdom network that captures intricate relationships between omic layers and entities.
• Cross-cohort Pattern Selection: Selects differential features and correlations based on robust patterns of fold change direction and correlation sign across cohorts.
• Topological Analysis: Evaluates network topology to identify key nodes that control subnetworks or mediate communication between kingdoms or subnetworks.
• Scalability: Handles millions of correlations required for network reconstruction and is designed for computational efficiency.
• Minimal Sample Size Requirement: Enables causal-inference analysis without requiring large sample sizes.

Scientific Applications:

• Host–microbiota interaction studies: Identifies candidate causal interactions between host and microbial features in microbiome research.
• Systems biology: Integrates multi-omic data to reveal system-level regulatory relationships and network architecture.
• Personalized medicine and disease mechanism discovery: Prioritizes putative causal drivers and regulatory nodes relevant to specific conditions or diseases.

Methodology:

Data integration of multi-omic data from multiple cohorts;
selection of differential features based on fold change direction and correlation sign across cohorts;
construction of a causal transkingdom network using causality-sensitive metrics, statistical thresholds, and topological criteria;
topological analysis to identify influential nodes within the reconstructed network.