AMOUNTAIN

AMOUNTAIN identifies active modules in multi-layer weighted gene co-expression networks (WGCNs) using a continuous optimization approach to discover functionally relevant connected subgraphs directly from expression profiles.

Key Features:

• Data-driven WGCN construction: Constructs weighted gene co-expression networks purely from expression profiles without requiring structural information, prior-knowledge databases, or external gene activity measures.
• Multi-layer network analysis: Models multiple layers (e.g., experimental conditions, time points, species) within a unified framework to detect modules across layers.
• Continuous optimization approach: Employs a continuous optimization technique to identify active connected subgraphs (active modules) within WGCNs.
• Improved precision over clustering methods: Focuses on relevant gene interactions and reduces inclusion of uninformative genes compared with clustering-based module detection.
• Validation on synthetic and real-world data: Assessed using both synthetic datasets and real-world expression data to evaluate robustness and biological relevance.

Scientific Applications:

• Systems biology and genomics: Identification of co-expression-based functional modules across conditions or datasets.
• Gene regulation and disease pathway discovery: Discovery of gene sets implicated in regulatory mechanisms and disease-related pathways.
• Comparative and temporal studies: Analysis of module dynamics or conservation across species, experimental conditions, or time points.
• Functional and evolutionary inference: Inference of functional organization of genes and evolutionary processes from module structure.

Methodology:

Constructs weighted gene co-expression networks from expression profiles; extends the model to multi-layer networks under a unified framework; applies continuous optimization to identify active connected subgraphs (active modules); validated on synthetic and real-world datasets.