Multiple protein structure alignment at scale with FoldMason

Area of Science:

• Structural bioinformatics
• Computational biology
• Protein structure analysis

Background:

• Protein structure conservation extends beyond sequence, necessitating multiple structural alignment (MSTA) for analyzing distantly related proteins.
• Advances in computational protein structure prediction generate vast datasets, demanding efficient and accurate MSTA methods.

Purpose of the Study:

• To introduce FoldMason, a novel progressive MSTA method designed for high-throughput analysis of large protein structure datasets.
• To evaluate FoldMason's speed and accuracy against existing state-of-the-art MSTA methods.

Main Methods:

• FoldMason employs a progressive alignment strategy, utilizing pairwise structural aligners Foldseek and TM-align.
• The method was tested on large-scale datasets, including Flaviviridae glycoproteins, to assess its performance.
• Confidence scores and interactive visualizations are integrated into the FoldMason workflow.

Main Results:

• FoldMason achieves alignment quality comparable to or exceeding current state-of-the-art MSTA methods.
• The method demonstrates a significant speed improvement, being two orders of magnitude faster than existing tools.
• FoldMason's multiple structural alignments facilitate robust phylogenetic analysis, extending beyond the 'twilight zone' of sequence similarity.

Conclusions:

• FoldMason provides essential speed and accuracy for large-scale protein structure analysis, crucial in the era of accurate structure prediction.
• The tool supports phylogenetic analysis of distantly related proteins, enhancing evolutionary studies.
• FoldMason is a free, open-source software, promoting accessibility for the scientific community.