Accelerate Your Discovery: Protein-Hunter is Now Live on Vecura

What is Protein-Hunter?

Protein-Hunter is a de novo structure-hallucination pipeline designed to invent new protein and cyclic peptide binders against specific targets from scratch. It leverages diffusion-based structure predictors—specifically Boltz1/2 or Chai1—in an iterative loop with sequence optimization tools like ProteinMPNN and LigandMPNN. By moving away from traditional scaffold-based design, it allows for the creation of binders against diverse targets, including proteins, small molecules, and nucleic acids.

It helps users rapidly explore the design space for novel therapeutic candidates or biological probes. It is especially useful for researchers targeting "hard-to-drug" surfaces or needing specialized cyclic peptides that offer improved metabolic stability compared to linear proteins.

What can users do with Protein-Hunter on Vecura?

With Protein-Hunter on Vecura, users can:

• Design de novo linear protein binders against a wide range of targets without needing a known scaffold.
• Create short macrocyclic peptides tailored to engage flat or featureless protein surfaces.
• Filter and rank candidates automatically using interface confidence scores (ipTM) and structural confidence (pLDDT).
• Perform optional cross-validation of generated designs using AlphaFold3 to improve candidate reliability.

What the output means

The output provides a comprehensive set of results, including atomic-coordinate files (CIF/PDB) of the designed complex, the optimized binder sequence, and detailed design statistics across multiple iterations. Additionally, it provides a summary CSV of all trajectories for deeper downstream analysis.

This output should be used to support scientific decision making. It does not replace experimental validation; users should treat these as in-silico candidates requiring physical synthesis and benchtop verification.

Why this matters

The ability to perform de novo design against arbitrary targets represents a major shift in biotechnology, moving from screening existing libraries to custom-inventing molecules tailored to specific molecular interfaces. By integrating cutting-edge diffusion models with automated optimization loops, Protein-Hunter lowers the barrier for designing binders against traditionally intractable targets.

This approach significantly accelerates the drug discovery pipeline by narrowing down the design space to high-confidence candidates before moving to resource-intensive laboratory synthesis. It democratizes access to sophisticated protein design workflows by automating the complex orchestration between diffusion predictors and sequence optimizers.

• Developed by: Yeh-Lin Cho et al.
• Source: Official GitHub Repository
• Reference: Protein-Hunter preprint (bioRxiv 2025)