Protein Pocket Detection with SiteFerret Now Available on Vecura

What is SiteFerret?

SiteFerret is a sophisticated hierarchical clustering tool designed to identify and rank ligand-binding pockets on protein surfaces. Starting from a PQR structure file, it uses the NanoShaper engine to analyze concave regions at multiple probe radii and applies a pre-trained Isolation Forest anomaly detector to assess the druggability of the detected sites. It helps users streamline the identification of protein binding sites, facilitating more efficient structure-based drug discovery. It is especially useful for researchers who need to prioritize candidate pockets for downstream molecular docking.

What can users do with SiteFerret on Vecura?

With SiteFerret on Vecura, users can:

• Automatically identify and rank potential ligand-binding pockets from a PQR protein structure file.
• Filter candidate sites using customizable druggability thresholds to refine results.
• Obtain detailed structural data, including pocket lining residues, PQR probe geometry, and surface triangulation files for visualization.
• Prepare protein structure data for downstream docking software like AutoDock Vina or Glide using the provided atom-level output.

What the output means

The output provides a ranked list of predicted binding pockets, where lower anomaly scores indicate higher predicted druggability. The results include specific residue listings, probe-sphere geometries, and surface files that can be loaded into tools like VMD or PyMOL.

This output should be used to support scientific decision-making. It does not replace experimental validation.

Why this matters

The accurate identification and prioritization of binding pockets are fundamental requirements for structure-based drug design. SiteFerret improves upon traditional methods by integrating multi-scale probe radius exploration with machine-learning-based scoring, allowing for a more nuanced understanding of protein surface topography.

By automating these complex computational steps, SiteFerret reduces the technical barriers to high-quality pocket analysis, enabling researchers to focus on hypothesis generation and experimental validation of therapeutic targets.

• Developed by: Concept Lab
• Source: Official GitHub Repository
• Reference: SiteFerret: beyond simple pocket identification in proteins (arXiv:2212.11888)