Smolign: a spatial motifs-based protein multiple structural alignment method.

Area of Science:

• Structural Bioinformatics
• Computational Biology
• Biochemistry

Background:

• Effective protein multiple structural alignment (MSTA) is crucial for identifying biologically significant structural motifs.
• Current MSTA methods often rely on pairwise comparisons, leading to suboptimal alignments, particularly for proteins with low similarity.
• Accurate MSTA aids in functional annotation and drug design.

Purpose of the Study:

• To introduce a novel strategy for protein multiple structural alignment (MSTA).
• To develop a sensitive and robust automated algorithm for MSTA.
• To improve the detection of similarities among protein structures, even under low similarity conditions.

Main Methods:

• A contact-window based motif library is constructed from protein structural data.
• Common alignment seeds are discovered and extended from this library.
• Optimal superimposition of multiple structures is achieved using an enhanced partial order curve comparison method based on these seeds.

Main Results:

• The novel strategy enables simultaneous detection of multiple correspondences and global alignment capture.
• The method supports flexible alignments, offering enhanced sensitivity and robustness.
• The new approach outperforms popular MSTA methods on diverse protein structure datasets with varying similarity levels.

Conclusions:

• The developed MSTA strategy provides superior alignment results compared to existing methods.
• The algorithm effectively identifies structural similarities in proteins, even with low sequence identity.
• A web-based tool and executable are available for broader application and validation.