urms

urms aligns local regions of 3D protein structures in PDB format using a hybrid Unit-Vector Root Mean Squared (URMS) and Root Mean Square (RMS) metric algorithm to detect and score structural similarities.

Key Features:

• Hybrid Algorithm Approach: Combines the URMS and RMS metrics, where URMS initially identifies potential rotations and an RMS-based dynamic programming algorithm refines those alignments.
• Efficient Transformation Space Search: Implements a fast screening protocol to navigate the transformation space, cluster initial rotations, and prioritize maximal local similarities.
• Detection of Complex Similarities: Identifies intricate structural similarities and repeats within protein domains by leveraging complementary strengths of URMS and RMS.
• Symmetric Results: Produces symmetric alignment outputs that are consistent regardless of the order of input structures.
• Statistical Significance Estimation: Includes a model that estimates statistical significance of alignments by considering both match score and RMS values.

Scientific Applications:

• Protein domain comparison: Local alignment of 3D structures to detect domain-level similarities.
• Evolutionary relationship inference: Exploring evolutionary relationships between proteins based on structural similarity.
• Functional motif identification: Identifying functional motifs through detected structural correspondences.
• Structural repeat analysis: Investigating repeats and complex architecture within protein families.
• Benchmarking and validation: Method evaluation and benchmarking using the SCOP (Structural Classification of Proteins) database.

Methodology:

Initial screening uses the URMS metric to propose potential rotations which are clustered, and an RMS-based dynamic programming algorithm is applied to refine clusters and identify maximal local similarities.