pandaR

pandaR integrates protein-protein interactions, gene expression profiles, and sequence motifs using a message-passing model (PANDA — Passing Attributes between Networks for Data Assimilation) to reconstruct condition-specific regulatory networks.

Key Features:

• PANDA algorithm: Implements a message-passing model for Passing Attributes between Networks for Data Assimilation (PANDA) to combine information across networks.
• Data integration: Assimilates protein-protein interactions, gene expression profiles, and sequence motifs as complementary data sources.
• Genome-wide prediction: Predicts regulatory relationships across the genome rather than limited loci.
• Condition-specific networks: Reconstructs regulatory networks specific to tissues, cell types, or experimental conditions.
• Improved accuracy: Integrates multiple datasets to improve reconstruction accuracy compared with single-dataset methods.
• Biological insight: Reveals specific biological mechanisms and pathways that may be missed by single-source approaches.
• Cross-organism applicability: Applicable to yeast and higher eukaryotes for studies across model organisms and complex organisms.
• Generalizable framework: Adaptable to incorporate various forms of genome-scale data, including additional regulatory interactions and expression profiles.

Scientific Applications:

• Regulatory network reconstruction: Reconstruction of genome-scale transcriptional regulatory networks from integrated datasets.
• Condition- and tissue-specific analysis: Derivation of regulatory networks specific to tissues, cell types, or experimental conditions.
• Pathway and mechanism discovery: Identification of regulatory pathways and mechanisms that are not apparent from single data types.
• Comparative regulatory studies: Comparative analysis of regulatory architecture across yeast and higher eukaryotes.

Methodology:

Uses a message-passing model to assimilate protein-protein interactions, gene expression profiles, and sequence motifs to predict regulatory relationships and reconstruct condition-specific regulatory networks.