> ## Documentation Index > Fetch the complete documentation index at: https://proto.evodesign.org/docs/llms.txt > Use this file to discover all available pages before exploring further. # MPNN Structure-Conditioned Mutation > LigandMPNN/ProteinMPNN-guided mutation of an existing protein sequence
MPNN Structure-Conditioned Mutation
**License:** This generator can use multiple tools, each under its own license. See the **Tools Used** tab and each tool's page for license details.

This generator is open source. Any third-party models, product names, or trademarks referenced are the property of their respective owners, and Proto is not affiliated with them.

evo-design/proto-language/proto\_language/generator/mpnn\_mutation\_generator.py
View source
Generator contributors
Mutate protein sequences using MPNN structure-conditioned probabilities. ## API Reference
ConfigMPNNMutationGeneratorConfig Source
Configuration for structure-conditioned MPNN mutation. The generator scores the current sequence against a backbone, chooses mutable positions using the model's probability of the current residue, then replaces each chosen residue from the model's per-position amino-acid distribution. Structure-conditioned model used for mutation probabilities: ligandmpnn or proteinmpnn. Options: `ligandmpnn`, `proteinmpnn` Use configured structure\_inputs or each proposal's attached structure for MPNN scoring. Options: `configured_structure_inputs`, `proposal_structure` Structures for MPNN scoring; templates when structure\_source='proposal\_structure'. Structure chain corresponding to the assigned sequence. Required for ambiguous multi-chain inputs. Number of positions to resample per sequence. Optional per-chain 1-indexed positions eligible for mutation. If unset, the output chain is mutable. Single-letter amino acids to forbid as replacement residues. 'sample' draws from MPNN probabilities; 'argmax' chooses the highest-probability residue. Options: `sample`, `argmax` Temperature applied to MPNN logits before replacement sampling. ProteinMPNN weights used when model='proteinmpnn'. Options: `proteinmpnn`, `v_48_002`, `v_48_010`, `v_48_030`, `abmpnn`, `soluble` Whether LigandMPNN scoring conditions on fixed-residue sidechain atoms. Ligand-residue distance cutoff (Å) used by LigandMPNN scoring. LigandMPNN implementation used for scoring: Foundry-backed ligand\_mpnn or original LigandMPNN. Options: `ligand_mpnn`, `original` Optional explicit LigandMPNN checkpoint path. Optional seed passed directly to LigandMPNN scoring; None uses Proto's derived seed stream. Use Proto's per-proposal derived seed or NumPy's global RNG stream for mutation sampling. Options: `derived_seed`, `global` Optional seed applied once when rng\_mode='global'. Optional MPNN scoring mode used to refresh pmpnn after applying mutations. Options: `disabled`, `single_aa`, `autoregressive` Optional structure preparation used to attach sequence-consistent structures after mutation. Device for MPNN scoring. Whether to print status messages during MPNN scoring.
## Usage ```python python icon="python" theme={null} from proto_language.generator import MPNNMutationGenerator, MPNNMutationGeneratorConfig from proto_language.core import Segment config = MPNNMutationGeneratorConfig( # Configure parameters here ) generator = MPNNMutationGenerator(config) segment = Segment(length=100, sequence_type="protein") generator.assign(segment) generator.sample() ``` ## Metadata | Property | Value | | ------------------------ | ----------------------- | | Key | `mpnn-mutation` | | Class | `MPNNMutationGenerator` | | Category | `mutation` | | Input Type | `starting_sequence` | | Uses GPU | `True` | | Supported Sequence Types | `protein` | | Allows Empty Start | `False` |