proteinmpnn-nim

Design protein sequences for PDB backbones using ProteinMPNN via NVIDIA NIM.

413|62|Updated Jun 23, 2026

Search Tags:#bioinformatics #protein design #nim #biochemistry #proteinmpnn #inverse folding

Ready to run

npx skills add https://github.com/NVIDIA-BioNeMo/bionemo-agent-toolkit --skill proteinmpnn-nim

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: proteinmpnn-nim
Source: https://github.com/NVIDIA-BioNeMo/bionemo-agent-toolkit/tree/main/plugins/bionemo-agent-toolkit/skills/proteinmpnn-nim
Command: npx skills add https://github.com/NVIDIA-BioNeMo/bionemo-agent-toolkit --skill proteinmpnn-nim

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, and includes references (resource) components.

What problem does it solve?

This Skill solves the challenge of inverse protein folding, allowing users to generate stable amino acid sequences that are compatible with a specific 3D protein backbone structure.

Core Features & Use Cases

Inverse Folding: Generate sequences for fixed backbones using NVIDIA NIM-accelerated ProteinMPNN.
Design Control: Fine-tune designs with sampling temperatures, amino acid exclusions, and chain-specific constraints.
Use Case: A researcher has a novel protein backbone design and needs to generate 50 candidate sequences to test for stability and binding affinity in a wet lab.

Quick Start

Use the proteinmpnn-nim skill to design 10 sequences for the backbone provided in the file 1R42.pdb using a sampling temperature of 0.1.

Frequently Asked Questions about proteinmpnn-nim

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I design protein sequences for a specific 3D backbone structure?▼

To design protein sequences for a specific 3D backbone, you can use inverse protein folding via the ProteinMPNN model. Provide a valid PDB file of the target backbone to generate compatible amino acid sequences.

What is inverse protein folding and when do I need it for biochemistry research?▼

Inverse protein folding generates stable amino acid sequences matching a fixed 3D protein backbone. It is needed when you have a novel backbone structure and want to produce candidate sequences for wet lab stability and binding affinity testing.

Can I fix certain amino acid positions or exclude specific residues during protein design?▼

Yes, you can fix certain amino acid positions and exclude specific residues during protein design. The model supports complex design constraints including fixed positions, amino acid exclusions, and multi-chain redesign tasks.

How do I generate multiple candidate sequences for wet lab testing from a PDB file?▼

Generate multiple candidate sequences from a PDB file by specifying the desired number of sequences and a sampling temperature. The tool outputs multi-FASTA files containing the sequences and associated confidence scores for downstream validation.

Does the ProteinMPNN model support multi-chain protein complex redesign?▼

Yes, the ProteinMPNN model supports multi-chain protein complex redesign. You can apply chain-specific constraints and design constraints to generate sequences for complex multi-chain backbone structures.

What format do I need for my protein backbone input and what output will I receive?▼

You need a valid PDB file for your protein backbone input. You will receive multi-FASTA outputs containing the designed sequences along with associated confidence scores for downstream structural validation.