GASH: an improved algorithm for maximizing the number of equivalent residues between two protein structures

Area of Science:

• Structural bioinformatics
• Computational biology
• Protein structure analysis

Background:

• Introduces GASH, a novel, publicly accessible program for structural alignment and superposition.
• Utilizes Number of Equivalent Residues (NER) for quantitative structural similarity assessment.
• Employs a genetic algorithm with conjugate gradient maximization for optimizing multiple alignments.

Purpose of the Study:

• To evaluate the performance of GASH against established structural alignment tools.
• To assess the accuracy and completeness of alignments generated by GASH.
• To provide a detailed analysis of GASH's capabilities in structural superposition.

Main Methods:

• Comparison of GASH with DaliLite, CE, and Global ASH on extensive test datasets.
• Evaluation of alignment quality using NER and residue counts under varying RMSD cutoffs.
• Detailed analysis of alignments for biologically relevant structural pairs.

Main Results:

• GASH demonstrates superior overall performance in superposition accuracy compared to DaliLite, Global ASH, and CE.
• Alignment crossover and initial alignment set completeness were examined.
• Numerical and detailed qualitative assessments of alignment quality were performed.

Conclusions:

• GASH achieves the best performance in structural superposition accuracy across different evaluation criteria.
• For proteins under 500 residues, GASH, DaliLite, and CE show similar CPU usage; DaliLite is faster for larger proteins.
• GASH is accessible via HTTP and SOAP interfaces at http://www.pdbj.org/GASH/.