RADAR (EBI)

RADAR identifies gapped approximate repeats and complex repeat architectures within protein sequences to characterize internal duplications and repeat unit organization.

Key Features:

• Automatic Segmentation: Implements a three-step algorithmic approach to segment query protein sequences into repeat units without prior assumptions about repeat number or length.
• Repeat Length Determination: Utilizes spacing between suboptimal self-alignment traces to ascertain repeat lengths.
• Optimization of Repeat Borders: Adjusts repeat borders to maximize an integer number of repeats and refine segmentation boundaries.
• Validation of Distant Repeats: Employs iterative profile alignment to validate and detect distant or divergent repeats.
• Comprehensive Detection: Detects short composition-biased repeats, gapped approximate repeats, and complex repeat architectures involving multiple repeat types.
• Autonomous Operation: Operates without manual intervention or prior knowledge of repeat structures, performing automated repeat detection and segmentation.
• Coverage and Accuracy Assessment: Demonstrates good coverage, accurate alignments, and reasonable repeat borders through comparative analysis with the Pfam-A database.
• Novel Repeat Discovery: Screening against databases such as Swissprot revealed approximately 3,000 repeats not previously annotated in existing domain databases.

Scientific Applications:

• Protein evolution analysis: Enables analysis of internal duplication and repeat-driven evolutionary processes in proteins.
• Functional and structural unit identification: Facilitates identification of repeat-derived functional and structural units within large proteins and aids annotation of novel repeats.

Methodology:

Analyzing self-alignment traces to determine repeat lengths; iterative profile alignment to refine and validate repeats; and comparison against databases such as Pfam-A and Swissprot for validation and discovery.