MULCCH

MULCCH applies multi-task spectral consensus clustering to identify shared and pathogen-specific host transcriptional regulatory modules across infections.

Key Features:

• Multi-task Spectral Clustering Algorithm: Extends a multi-task clustering algorithm to handle hierarchically related tasks and complex datasets measuring host transcriptional responses across different infections.
• Consensus Extension: Incorporates a consensus approach to enhance the robustness and accuracy of identifying gene expression modules and detecting genes that change module membership between infections.
• High-confidence Strain-specific Inference: Infers high-confidence strain-specific host response modules across multiple virus strains.
• Large-scale Dataset Processing: Processes large collections of datasets that capture host transcriptional responses to various pathogens.
• Statistical Robustness: Employs advanced clustering techniques to ensure the statistical robustness of identified gene expression modules.
• Coherence and Enrichment in Applications: Produces clusters that demonstrate greater coherence and are enriched for immune system-related processes and regulators in analyses such as mammalian responses to influenza viruses.

Scientific Applications:

• Systems biology of infectious diseases: Facilitates comparison and identification of gene expression modules across different infections to pinpoint shared and strain-specific regulatory components.
• Comparative module analysis: Detects genes that alter module membership between infections to highlight pathogen-specific transcriptional changes.
• Pathway and gene-set characterization: Identifies molecular pathways and gene sets that characterize commonality and specificity of host responses to different viral pathogenicities.
• Influenza response analysis: Applied to mammalian transcriptional responses to influenza viruses to derive coherent, immune-enriched clusters and regulators.

Methodology:

Processes large collections of host transcriptional response datasets, employs advanced clustering techniques (multi-task spectral consensus clustering) to ensure statistical robustness of modules, and yields clusters that show increased coherence and enrichment for immune-related processes and regulators in practical applications.