Connectivity independent protein-structure alignment: a hierarchical approach

Area of Science:

• Structural bioinformatics
• Computational biology
• Biochemistry

Background:

• Protein structure alignment is crucial for understanding protein function, evolution, and model building.
• Existing methods often overlook proteins with similar spatial arrangements of secondary structure elements (SSEs) but different polypeptide chain connectivities (non-sequential SSE connectivity).

Purpose of the Study:

• To develop and evaluate a novel protein-structure alignment method, GANGSTA, capable of handling non-sequential SSE connectivity.
• To assess GANGSTA's ability to detect structural similarity and functional relationships in proteins with diverse chain connectivities.

Main Methods:

• A two-level hierarchical approach implemented in the GANGSTA program.
• Level 1: Genetic algorithm (GA) to maximize contacts and relative orientations between SSEs (alpha-helices and beta-strands).
• Level 2: Optimization of residue pair contacts based on the best SSE alignments.

Main Results:

• GANGSTA successfully performs protein-structure alignment, considering non-sequential SSE connectivity.
• The method detects significant structural similarity in functionally important folds, even with differing chain arrangements.
• Performance for sequential SSE connectivity is comparable to existing structure alignment methods.

Conclusions:

• GANGSTA effectively identifies meaningful protein-structure alignments irrespective of SSE connectivity.
• The program can uncover structural similarities between protein folds from different superfamilies that share similar structures and functions despite variations in SSE connectivity.