bionemo-agent-toolkit
Protein design, molecule generation, docking, and genomics workflows
All Skills in This Repository (33)
Pure Emerald Level Indicators
NVIDIA-BioNeMokermt-embed
Extract per-molecule embeddings from KERMT checkpoints for SMILES processing.
NVIDIA-BioNeMoopenfold2-nim
Predict monomer protein structures from amino-acid sequences via OpenFold2 NIM microservices.
NVIDIA-BioNeMocomplexa-evaluate-pdbs
Evaluate directories of PDB files by orchestrating refolding and structural analysis workflows.
NVIDIA-BioNeMoopenfold3-nim
Predict 3D biomolecular structures for proteins, DNA, RNA, and ligands via OpenFold3 NIM.
NVIDIA-BioNeMoboltz2-nim
Predict 3D biomolecular structures and ligand binding affinities for proteins, DNA, and RNA using NVIDIA NIM microservices.
NVIDIA-BioNeMomsa-structure-prediction-pipeline
Chain MSA-Search and OpenFold3 NIM microservices to predict protein structures.
NVIDIA-BioNeModiffdock-nim
Predict small-molecule binding poses against protein targets using DiffDock via NVIDIA NIM.
NVIDIA-BioNeMoparabricks
Route NVIDIA Parabricks pbrun tools with version-aware command guidance for genomics workflows.
NVIDIA-BioNeMocuequivariance
Construct group-equivariant neural network layers using irreducible representations and segmented tensor products.
NVIDIA-BioNeModrug-discovery-pipeline
Generate, dock, and score drug candidates via GenMol, DiffDock, and Boltz2 NIMs.
NVIDIA-BioNeMogenomics-workflow-acceleration
Map CPU-bound genomics workflow steps to NVIDIA Parabricks GPU equivalents.
NVIDIA-BioNeMogenmol-nim
Generate novel drug-like molecules via SAFE notation using GenMol NIM.
Frequently Asked Questions
FAQPage SchemaHow to install bionemo-agent-toolkit?▼
Run `npx skills add NVIDIA-BioNeMo/bionemo-agent-toolkit --all -g -y` in your terminal to install every skill globally. To add a single skill instead, use `npx skills add NVIDIA-BioNeMo/bionemo-agent-toolkit --skill boltz2-nim -g -y`.
What can I do with the BioNeMo skills?▼
You can predict protein and complex structures, design de novo protein binders, generate and optimize drug-like molecules, dock ligands, score binding affinity, and accelerate genomics pipelines. Each skill handles endpoint setup, request formatting, and result validation for you.
Do I need a GPU to use these skills?▼
Not always. Most skills call hosted NVIDIA API endpoints that only need an NGC API key, while local Docker deployment of models like Boltz2, Evo 2, or Proteina-Complexa requires a CUDA-capable NVIDIA GPU.
Which agents work with this toolkit?▼
All skills follow the standard SKILL.md format and run in Claude Code, Codex, Cursor, and other compatible agents. The repo also ships plugin marketplace files for native installation in Claude Code and Codex.
Can I run a full drug discovery pipeline?▼
Yes. The drug-discovery-pipeline skill chains molecule generation with GenMol, docking with DiffDock, and affinity prediction with Boltz2 into one automated workflow from a single request.
Related Repositories in Education & Research
View All in Education & Research→
shareAI-lablearn-claude-code
Learn to build AI agent harnesses from scratch
microsoftai-agents-for-beginners
Learn to build, deploy, and secure AI agents step by step
leonardomso33-js-concepts
Master 33 essential JavaScript concepts with tested, verified guides