E1DS

E1DS predicts enzyme catalytic sites and annotates enzyme sequences by mining long conserved sequence motifs to derive sequence signatures that identify catalytic residues.

Key Features:

• Pattern Mining Algorithm: Employs a pattern mining algorithm to discover long motifs composed of several sequential conserved blocks that are significantly conserved among proteins within the same Enzyme Commission (EC) group.
• Sequence Signatures: Derives sequence signatures composed of at least three sequentially conserved blocks that frequently co-occur in enzyme sequences, reflecting that catalytic sites often comprise residues distributed across different primary-structure regions.
• Extensive Database: Contains 5,421 sequence signatures covering 932 four-digit EC numbers.

Scientific Applications:

• Enzyme sequence annotation: Annotates enzyme sequences at scale by matching sequence signatures to enzyme families and EC groups.
• Catalytic site and residue identification: Predicts catalytic sites and involved residues with sensitivity reported to be higher than PROSITE, validated against the Catalytic Site Atlas.

Methodology:

Pattern discovery via a pattern mining approach to identify long motifs representing conserved blocks; construction of sequence signatures from these motifs requiring multiple sequential conserved blocks observed together; and evaluation/validation by comparison to the Catalytic Site Atlas demonstrating higher sensitivity relative to existing databases such as PROSITE.