oteam/solublempnn

ProteinMPNN

Pre-trained model for protein sequence design and scoring conditioned on a fixed backbone.

Disclaimer

This is an UNOFFICIAL implementation of the Robust deep learning-based protein sequence design using ProteinMPNN by Justas Dauparas, et al.

The OFFICIAL repository of ProteinMPNN is at dauparas/ProteinMPNN.

The team releasing ProteinMPNN did not write this model card for this model so this model card has been written by the OFoldX team.

Model Details

ProteinMPNN is an autoregressive inverse-folding model that designs an amino-acid sequence for a fixed protein backbone. It represents residues as a spatial nearest-neighbor graph, encodes backbone geometry with message-passing blocks, and decodes residues in a sampled causal order. Generation supports fixed positions, tied identities, position-specific amino-acid constraints, and sequence profiles.

Please refer to the Training Details section for more information on the training process.

Variants

• oteam/proteinmpnn: General-purpose ProteinMPNN using N, CA, C, and O backbone atoms.
• oteam/solublempnn: ProteinMPNN trained for soluble-protein sequence design.
• oteam/proteinmpnn-ca-only: Reduced model for backbones represented only by CA atoms.

Model Specification

Variant	Backbone Atoms	Hidden Size	Encoder Blocks	Decoder Blocks	Neighbors	Num Parameters (M)
Standard	N, CA, C, O	128	3	3	48	1.660
Soluble	N, CA, C, O	128	3	3	48	1.660
CA-only	CA	128	3	3	48	1.629

Links

• Checkpoint: v_48_020.pt
• Upstream license: LICENSE
• Upstream source revision: 8907e6671bfbfc92303b5f79c4b5e6ce47cdef57
• Code: ofoldx.models.proteinmpnn
• Paper: Robust deep learning-based protein sequence design using ProteinMPNN
• Developed by: Justas Dauparas, Ivan Anishchenko, Nathaniel Bennett, Hua Bai, Robert J. Ragotte, Lukas F. Milles, Basile I. M. Wicky, Alexis Courbet, Rob J. de Haas, Neil Bethel, Philip J. Y. Leung, Timothy F. Huddy, Sam Pellock, Doug Tischer, Frederick Chan, Brian Koepnick, Hannah Nguyen, Alex Kang, Banumathi Sankaran, Asim K. Bera, Neil P. King, and David Baker
• Model type: Autoregressive message-passing neural network for fixed-backbone protein sequence design
• Original Repository: dauparas/ProteinMPNN

Usage

The model file depends on the ofoldx library. You can install it using pip:

pip install ofoldx

Direct Use

Sequence Design

You can use this model directly with a pipeline for fixed-backbone protein sequence design:

import ofoldx
from ofoldx.io import read_structure

structure, _ = read_structure("backbone.cif")
pipeline = ofoldx.pipeline("design", model="oteam/solublempnn")
result = pipeline(structure)

Each candidate contains the designed molecular target and named sequence scores.

Sequence Scoring

You can use this model directly with a pipeline for fixed-backbone protein sequence scoring:

import ofoldx
from ofoldx.io import read_structure

structure, _ = read_structure("backbone.cif")
pipeline = ofoldx.pipeline("score", model="oteam/solublempnn")
result = pipeline(structure, num_decoding_orders=8, seed=7)

The result contains designed and global sequence negative log likelihood (NLL) in nats per residue.

Downstream Use

Design and Packing

Combine a designer and an independently loaded MPNN Packer with ProteinMpnnDesignAndPackPipeline to generate side-chain coordinates after sequence design. Both operations use the coordinate-observation policy selected during preparation.

See the pipeline guide for staged execution with prepare, forward, and postprocess.

Training Details

ProteinMPNN is trained to recover amino-acid identities from protein backbone geometry.

Training Data

ProteinMPNN was trained on fixed protein backbones paired with native amino-acid sequences. The soluble variant focuses on soluble proteins, while the CA-only variant uses reduced backbone geometry.

Training Procedure

ProteinMPNN uses autoregressive sequence recovery with coordinate perturbations for robustness.

OFoldX exposes differentiable encode and forward operations and the masked, label-smoothed MpnnSequenceLoss for training and fine-tuning. Dataset construction, parsing, clustering, and complete upstream training orchestration remain the training application's responsibility.

Citation

@article{dauparas2022robust,
  author = {Dauparas, Justas and Anishchenko, Ivan and Bennett, Nathaniel and others},
  title = {Robust deep learning-based protein sequence design using ProteinMPNN},
  journal = {Science},
  volume = {378},
  number = {6615},
  pages = {49--56},
  year = {2022},
  doi = {10.1126/science.add2187}
}

The artifacts distributed in this repository are part of the OFoldX project. If OFoldX supports your research, please cite the OFoldX project as follows:

@software{chen2026ofoldx,
  author = {Chen, Zhiyuan},
  title  = {OFoldX},
  year   = {2026},
  url    = {https://github.com/OTeam-AI4S/OFoldX},
}

Contact

Please use GitHub issues of OFoldX for any questions or comments on the model implementation or model card.

Please contact the authors of the ProteinMPNN paper for questions or comments on the paper/model.

License

The model weights, implementation, configuration, and documentation distributed by OFoldX are licensed under the GNU Affero General Public License.

For additional terms and clarifications, please refer to our License FAQ.

SPDX-License-Identifier: AGPL-3.0-or-later