NCutYX

NCutYX performs integrative clustering of multilayer omics data to identify groups of genes and proteins and sample clusters while revealing regulatory relationships across genomics, proteomics, and epigenetics layers.

Key Features:

• Multilayer Network Representation: Employs a multilayer network framework that captures dependencies and interactions across different biological layers.
• Variable Clustering: Identifies groups of variables (e.g., genes or proteins) that exhibit similar patterns across samples.
• Sample Clustering: Groups samples based on their omics profiles to identify distinct biological conditions or disease states.
• Biclustering: Simultaneously clusters both variables and samples to uncover subsets where specific variables behave similarly across a subset of samples.
• Comprehensive Analysis Pipeline: Implements normalization and integration steps that accommodate regulatory relationships in multilayer omics data.
• Novel Methodological Approaches: Introduces methods for handling the complexity of multilayer data, targeting analyses of complex diseases involving interacting biological layers.

Scientific Applications:

• Disease Mechanism Inference: Enables inference of disease mechanisms from integrated multilayer omics data.
• Biomarker Discovery: Supports identification of potential biomarkers across genomics, proteomics, and epigenetics data.
• Therapeutic Target Exploration: Facilitates exploration of therapeutic targets by linking inter-layer regulatory relationships.
• Multi-omics Analysis: Applicable to genomics, proteomics, and epigenetics studies for integrative analysis.

Methodology:

Uses the normalized cut technique to partition data by minimizing dissimilarity within clusters while maximizing dissimilarity between clusters, and includes normalization and integration steps that accommodate regulatory relationships across multilayer omics.