Use AlphaCutter Online

Commercially Available AlphaCutter No-Code Web Server

AlphaCutter: Removing Non-Globular Regions from Predicted Protein Structures

AlphaCutter is a tool for removing non-globular regions from predicted protein structures, leaving clean domain regions for downstream structural bioinformatics. It was developed by C. Tam and W. Iwasaki and published in Proteomics in 2023. The software is available on GitHub, and AlphaCutter-cleaned SwissProt structures are available on Zenodo.

A huge number of high-quality predicted protein structures are now publicly available. However, many of them contain non-globular regions, such as long loops and extended helices, which diminish the performance of downstream structural bioinformatic applications.

Confidence scores like pLDDT do not catch all of these regions. AlphaCutter was built to identify and remove them efficiently, so that predicted structures can be used more reliably for tasks such as protein redesign.

Attribute

Detail

Task

Clean non-globular regions from predicted structures

Publication

Proteomics, 2023

Input

Predicted protein structures in PDB format

Speed

Under 3 seconds per structure on average

Large-scale test

542,380 predicted SwissProt structures

How AlphaCutter Identifies Non-Globular Regions and Domains

According to the project documentation, AlphaCutter classifies segments as loops, helices or domain fragments using adjustable length cutoffs and contact information. Loops or helices that make non-local contacts are reclassified as domain fragments, and the cleaned domains can be written out together or separately.

  • Loop and helix thresholds: Minimum loop length (default 20) and minimum helix length (default 30) set what counts as non-globular.

  • Fragment and domain size: Minimum fragment length (default 5) and minimum domain length control which pieces are kept.

  • Confidence filter: An optional minimum average pLDDT excludes low-confidence domains.

  • Output format: All domains in a single PDB file, or each non-contacting domain as its own PDB file.

AlphaCutter Results: Cleaner Domains and Better Redesign Metrics

In a large-scale cleaning of 542,380 predicted SwissProt structures, the authors report that AlphaCutter removes non-globular regions that cannot be detected using pLDDT scores while preserving high integrity of the cleaned domain regions. As an application, it improved folding energy scores and sequence recovery rates in the redesign of domain regions.

What is Tamarind Bio?

Tamarind Bio is a no-code bioinformatics platform built to give life scientists and researchers access to powerful computational tools. Many cutting-edge machine learning models are hard to deploy and use. Tamarind provides an intuitive, web-based environment that removes the complexity of high-performance computing, software dependencies and command-line interfaces.

The platform is designed for biologists, chemists and other researchers who may not have a background in programming or cloud infrastructure but want to run models on their own data. Key features include:

  • A user-friendly graphical interface for setting up and launching experiments

  • A robust API for integration into existing research pipelines

  • An automated system for managing and scaling computational resources

Tamarind treats information and data security as a top priority, as detailed in its Trust Center and Terms of Service.

Accelerating Discovery with AlphaCutter on Tamarind Bio

  • Structure preprocessing: Clean predicted structures before running downstream structural analysis.

  • Protein redesign: Improve inputs for domain redesign, where the authors saw better folding energy scores and sequence recovery.

  • Large-scale curation: Process many predicted structures quickly, at under 3 seconds each on average.

  • Domain extraction: Obtain individual non-contacting domains as separate structures.

How to Use AlphaCutter on Tamarind Bio

  1. Log in: Go to tamarind.bio and open the AlphaCutter tool.

  2. Upload structures: Provide predicted protein structures in PDB format.

  3. Set thresholds: If exposed, adjust the minimum loop and helix lengths, fragment and domain minimums, and the pLDDT cutoff.

  4. Choose output: If exposed, pick a single combined PDB or one PDB per domain.

  5. Run: Submit the job.

  6. Download: Retrieve the cleaned structures.

  7. Inspect: View the cleaned domains against the originals to confirm the right regions were removed.

  8. Use downstream: Feed cleaned domains into redesign, docking or other structure-based workflows.

Things to Keep in Mind

  • Default thresholds may not suit every protein; the documentation suggests lowering loop and helix cutoffs to remove shorter segments.

  • The documentation advises against changing the local contact range, and retesting the loop and helix cutoffs if you do.

  • The tool operates on predicted structures; results depend on the quality of the input models.

Source: Tam C., Iwasaki W. AlphaCutter: Efficient removal of non-globular regions from predicted protein structures. Proteomics (2023). DOI: 10.1002/pmic.202300176. Code: https://github.com/johnnytam100/AlphaCutter. Cleaned SwissProt structures: https://doi.org/10.5281/zenodo.7944483

Supporting 10,000+ scientists around the world,

from leading biotechs, and global biopharma