Multiple protein structure alignment at scale with FoldMason
This article has been Reviewed by the following groups
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
- Evaluated articles (Arcadia Science)
Abstract
Protein structure is conserved beyond sequence, making multiple structural alignment (MSTA) essential for analyzing distantly related proteins. Computational prediction methods have vastly extended our repository of available protein structures, requiring fast and accurate MSTA methods. We introduce FoldMason, a progressive MSTA method that leverages the pairwise structural aligners Foldseek and TM-align for the multiple alignment of hundreds of thousands of protein structures, matching or exceeding the alignment quality of state-of-the-art MSTA methods while being two orders of magnitude faster. Using Flaviviridae glycoproteins, we demonstrate that FoldMason’s MSTAs support phylogenetic analysis beyond the “twilight zone.” FoldMason computes confidence scores, offers interactive visualizations, and provides essential speed and accuracy for large-scale protein structure analysis in the era of accurate structure prediction. FoldMason is free, open-source software.
Article activity feed
-
-
For future work, we plan to integrate the ProstT5 proteinlanguage model (37) to directly predict 3Di from aminoacid sequences, eliminating the need for slow structurepredictions. This integration will accelerate input generationfor FoldMason by over 3000× compared to optimizedColabFold prediction. Instead of structure input, an AAFASTA file can be provided for sequence-based MSTA.This approach would be particularly beneficial for studiesinvolving long proteins, as is the case in Mifsud et al
This is a really exciting prospect! I wasn't 100% convinced given the marginal improvements to LDDT-vs-time presented, but with this speed increase I would be all-in on FoldMason. Looking forward to reading this, and thanks for your work so far!
-
Caratta
I think this is meant to be "Caretta-shape."
-
