CD-Search

CD-Search identifies conserved structural and functional domains in protein sequences to support annotation, evolutionary analysis, and structural interpretation.

Key Features:

• Domain Detection: CD-Search employs RPS-BLAST and Position Specific Score Matrices (PSSMs) derived from Conserved Domain Database (CDD) alignments to identify conserved domains by aligning queries to domain-model consensus sequences.
• BLAST heuristics: It leverages BLAST heuristics to perform fast searches through the collection of domain models in the CDD.
• Visualization: Results include domain architecture cartoons, pairwise alignments between query and domain models, multiple alignment displays, and links to three-dimensional molecular graphics of known structures within domain families.
• Integration with Entrez: The CDD is indexed as a standalone database within the Entrez system, enabling cross-references to other Entrez resources such as MEDLINE and sequence records.
• Comprehensive Domain Models: The CDD mirrors public collections such as SMART and PFAM and includes NCBI-curated models structured to reflect conserved core substructures and preserved functional sites among family members.
• Automatic BLAST searches: CD-Search is run automatically by default for protein-protein queries submitted to the BLAST service.

Scientific Applications:

• Domain annotation: Identification of conserved domains to annotate protein sequences and locate preserved functional sites.
• Function prediction: Inferring protein function from detected domain architectures and conserved motifs.
• Evolutionary analysis: Assessing evolutionary relationships and domain architecture conservation among protein families.
• Structural interpretation: Linking domain models to known three-dimensional structures to support structural biology studies.
• Cross-database research integration: Combining domain annotations with Entrez and MEDLINE records to connect sequence features with literature and other database entries.

Methodology:

CD-Search uses RPS-BLAST with PSSMs derived from CDD alignments and BLAST heuristics to align query protein sequences to domain-model consensus sequences; the CDD is indexed within Entrez and CD-Search is run by default for protein–protein BLAST queries.