name: proteinmpnn
description: >
Design protein sequences using ProteinMPNN inverse folding. Use this skill when:
(1) Designing sequences for generated backbones (e.g. from BindCraft),
(2) Redesigning existing protein sequences,
(3) Fixing specific residues while designing others,
(4) Optimizing sequences for expression or stability,
(5) Multi-state or negative design.
For backbone generation, use boltzgen (recommended) or bindcraft.
For ligand-aware design, use ligandmpnn.
For solubility optimization, use solublempnn.
license: MIT
category: design-tools
tags: [sequence-design, inverse-folding]
source: https://github.com/adaptyvbio/protein-design-skills
ProteinMPNN Sequence Design
Prerequisites
| Requirement |
Minimum |
Recommended |
| Python |
3.8+ |
3.10 |
| CUDA |
11.0+ |
11.7+ |
| GPU VRAM |
8GB |
16GB (T4) |
| RAM |
8GB |
16GB |
How to run
Option 1: Local installation (recommended)
git clone https://github.com/dauparas/ProteinMPNN.git
cd ProteinMPNN
python protein_mpnn_run.py \
--pdb_path backbone.pdb \
--out_folder output/ \
--num_seq_per_target 16 \
--sampling_temp "0.1"
Option 2: Modal (via LigandMPNN wrapper)
cd biomodals
modal run modal_ligandmpnn.py \
--pdb-path backbone.pdb \
--num-seq-per-target 16
Key parameters
| Parameter |
Default |
Range |
Description |
--pdb_path |
required |
path |
Single PDB input |
--num_seq_per_target |
1 |
1-1000 |
Sequences per structure |
--sampling_temp |
"0.1" |
"0.0001-1.0" |
Temperature (must be string!) |
--pdb_path_chains |
all |
A,B |
Chains to design |
Temperature guide
| Temp |
Effect |
| 0.1 |
Low diversity, high recovery (production) |
| 0.2 |
Moderate diversity (default) |
| 0.3 |
Higher diversity (exploration) |
| 0.5+ |
Very diverse, lower quality |
Common mistakes
✅ --sampling_temp "0.1" — String with quotes
❌ --sampling_temp 0.1 — Float without quotes may cause errors
✅ --pdb_path_chains A,B — No spaces
❌ --pdb_path_chains A, B — Space after comma
Variants Comparison
| Variant |
Use Case |
| ProteinMPNN |
General |
| SolubleMPNN |
Expression (E. coli) |
| LigandMPNN |
Small molecules/metals |
Output format
output/
├── seqs/backbone.fa # FASTA sequences
└── backbone_pdb/backbone_0001.pdb
Typical performance
| Campaign |
Time (T4) |
Cost (Modal) |
| 100 × 8 seq |
15-20 min |
~$2 |
| 500 × 8 seq |
1-1.5h |
~$8 |
Throughput: ~50-100 sequences/minute.
Troubleshooting
| Error |
Cause |
Fix |
CUDA out of memory |
Long protein |
Reduce batch_size |
KeyError: 'A' |
Chain not in PDB |
Check chain IDs |
JSONDecodeError |
Invalid JSONL |
Validate JSON syntax |
Next: Structure prediction → protein-qc for filtering.
By