FiRES

FiRES detects repeated structural elements in protein structures de novo using topology-independent structure alignment to identify repeating motifs independent of sequence similarity.

Key Features:

• De novo repeat detection: Identifies proteins with repeated structural elements without relying on sequence similarity.
• Topology-independent alignment: Employs a topology-independent structure alignment method to detect repeating motifs within protein structures.
• Low sequence identity sensitivity: Effective at detecting repeats even when residue identity is less than 20%.
• Evolutionary signal detection: Uncovers structurally similar elements that may arise from gene duplication and fusion events or from convergent evolution driven by physicochemical constraints.
• Benchmark datasets: Evaluated against curated repeat databases MALIDUP (very divergent duplicated domains) and RepeatsDB (short tandem repeats).
• Comparative benchmarking: Performance benchmarked alongside lalign, RADAR, HHrepID, CE-symm, ReUPred, and Swelfe.
• Accuracy metrics: Reported accuracy = 0.86 for detecting proteins with duplicated domains and accuracy = 0.92 for proteins containing multiple repeated units.

Scientific Applications:

• Protein domain evolution: Enables detection of duplicated and rearranged domains to study evolutionary processes of protein domains.
• Structure classification: Supports refinement of structure classification by identifying recurring structural motifs across divergent sequences.
• Structure prediction: Provides structural repeat information that can inform and improve structure prediction algorithms.
• Protein engineering: Reveals repeat motifs and modular architectures useful for rational design and engineering of proteins.

Methodology:

Uses a topology-independent structure alignment approach for de novo detection of repeating motifs, with performance evaluated by benchmarking against MALIDUP and RepeatsDB and comparisons to lalign, RADAR, HHrepID, CE-symm, ReUPred, and Swelfe.