lcmix

lcmix integrates multiple genomic datasets as an R package using hierarchical graphical mixture models to identify genes and cis-regulatory regions involved in gene regulation and expression by combining transcription factor binding measurements, gene expression profiles, and sequence conservation metrics.

Key Features:

• Hierarchical graphical mixture models: Implements hierarchical and graphical mixture model approaches for integrating heterogeneous genomic data.
• Supported data types: Integrates transcription factor binding measurements, gene expression profiles, and sequence conservation metrics.
• Heterogeneous data integration: Combines datasets with varying biological meanings and statistical distributions into a unified framework.
• Regulatory element and gene identification: Identifies genes and cis-regulatory regions that play specific roles in biological pathways.
• Model topology evaluation: Evaluates different model topologies to optimize data integration.
• Biological and statistical interpretation: Assesses model effectiveness through biological and statistical interpretation.
• Computational efficiency: Enables rapid model fitting through computationally efficient implementations.
• Demonstrated use cases: Applied to Hedgehog and Dorsal signaling pathways in Drosophila to illustrate analysis of embryonic development.

Scientific Applications:

• Regulatory region and gene discovery: Identify key regulatory regions and genes involved in development and disease.
• Pathway-specific analysis: Analyze signaling pathways such as Hedgehog and Dorsal in Drosophila to study embryonic development.
• Integrative analysis of gene regulation: Provide a comprehensive understanding of gene regulation and expression patterns by integrating heterogeneous datasets.
• Model-driven biological insight: Use model comparison and evaluation to derive biological and statistical insights.

Methodology:

Uses hierarchical and graphical mixture models to integrate transcription factor binding, gene expression, and sequence conservation data, evaluates alternative model topologies, and performs computationally efficient model fitting.