PRODOC

PRODOC compares proteins by representing them as sequences of functional domains to identify functional relationships, domain rearrangements (including circular permutations), and gene fusion events for evolutionary and comparative genomics analyses.

Key Features:

• Domain-Based Protein Comparison: Represents proteins as domain sequences rather than amino acid sequences to identify proteins with similar functions despite differing domain arrangements, including jumbled or circularly permuted domains.
• Querying and Identification: Accepts queries as sequences of domains to identify proteins with identical, rearranged, or circularly permuted domain architectures, enabling functional hypothesis generation for multi-domain proteins.
• Alignment Generation: Generates alignments between multi-domain proteins based on their domain sequences to reveal evolutionary relationships and functional similarities.
• Superfamily Information: Provides information about distantly related domain families that form superfamilies to support evolutionary inference.
• Domain Sharing and Gene Fusion Analysis: Identifies domain sharing and gene fusion events across organisms to investigate genomic evolution and functional diversification.
• Genome-Genome Comparison: Performs exhaustive genome-to-genome comparisons to detect successive domain sharing and fusion events and to identify conserved gene clusters involved in similar biological processes among related species.

Scientific Applications:

• Protein Function Inference: Infers functions of multi-domain proteins by matching domain architectures rather than relying solely on primary sequence homology.
• Evolutionary Analysis: Analyzes evolutionary relationships and superfamily-level connections through detection of domain rearrangements, sharing, and fusion events.
• Comparative Genomics and Genomic Architecture: Investigates genomic architecture and conserved biological processes by detecting domain sharing, gene fusions, and conserved gene clusters across genomes.

Methodology:

Domain assignment is performed using hidden Markov models (HMMs) to represent proteins as sequences of functional domains for domain-centric comparison and alignment generation.