NVIDIA-Digital-Bio

NVIDIA-Digital-Bio / Proteina-Complexa

Public

Generative model for protein binder design for protein and small molecule targets. Combines a pretrained flow-based generative model (built on La-Proteina) with inference-time optimization.

12
0
100% credibility
Found Mar 16, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Proteina-Complexa is a generative AI model from NVIDIA for designing protein binders and scaffolds that bind to target proteins or small molecules.

How It Works

1
🧬 Discover Proteina-Complexa

You hear about this tool from NVIDIA that helps design custom proteins to bind to targets, perfect for drug discovery or biology research.

2
🛠️ Set up your workspace

Run a simple script or Docker command to prepare everything, like setting up a new kitchen for cooking.

3
📁 Load your target protein

Pick a protein structure file from your data or examples, and tell the tool what to bind to.

4
🚀 Launch binder design

Click run with your settings, and watch as it generates new protein binders tailored to your target.

5
👀 Review your designs

See the new proteins visualized, check scores, and filter the best ones.

🎉 Success: Binders ready!

You now have experimentally validated protein designs to test in the lab or advance your research.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 12 to 12 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is Proteina-Complexa?

Proteina-Complexa is a Python-based generative AI tool on GitHub for designing protein binders that target proteins or small molecules. It uses flow matching—a generative modeling technique that estimates gradients of the data distribution—to create atomistic structures, then refines them via inference-time search with rewards from tools like AlphaFold2 or RoseTTAFold3. Developers get CLI commands like `complexa design` to generate, filter, evaluate, and analyze designs, outputting validated PDBs with metrics like binder refolding success and interface quality.

Why is it gaining traction?

It unifies generative modeling via drifting with test-time optimization, beating prior methods on benchmarks for success rates and compute efficiency—key for real-world protein engineering. Wet-lab validation confirms in-silico hits bind targets, and extensions handle ligands, motifs, and enzymes. The NGC-hosted models and Docker/UV setup make it plug-and-play for generative models GitHub workflows.

Who should use this?

Protein engineers designing therapeutic binders for drug discovery targets like PD-L1. Computational biologists scaffolding motifs around ligands or validating de novo proteins. Structural biologists needing sequence co-design with ProteinMPNN or LigandMPNN alongside structure prediction.

Verdict

Strong pick for cutting-edge generative protein design, backed by an ICLR 2026 oral paper and solid docs/CLI/tests—despite 12 stars and 1.0% credibility signaling early maturity. Download models and run if you're in biomolecular modeling; expect tweaks as community grows.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.