Atomic context-conditioned protein sequence design using LigandMPNN
Peer-reviewed| Поле | Значение |
|---|---|
| Authors | Justas Dauparas, Gyu Rie Lee, Robert Pecoraro, Linna An, Ivan Anishchenko, Cameron Glasscock, David Baker |
| Year | 2025 |
| Venue / status | Peer-reviewed journal; Nature Methods (Q1) |
| DOI | 10.1038/s41592-025-02626-1 |
| URL | https://doi.org/10.1038/s41592-025-02626-1 |
| Preprint URL | https://doi.org/10.1101/2023.12.22.573103 |
| Tools | LigandMPNN |
Задача и claimed novelty
| Поле | Значение |
|---|---|
| Main biological task | Atomic-context inverse folding around small molecules, nucleotides and metals |
| Claimed novelty | NR |
Архитектура
| Поле | Значение |
|---|---|
| Model architecture | ProteinMPNN-family graph model with non-protein atomic context and side-chain packing |
| Input representation | NR |
| Encoder / decoder | NR |
| Generative state | Sequence + side-chain conformations |
| Conditioning | Design masks; fixed residues; non-protein atoms; chain context |
| Sampling | NR |
Training
| Поле | Значение |
|---|---|
| Objectives / losses | NR |
| Dataset | NR |
| Preprocessing | NR |
Inference
NR
Baselines и retrospective benchmarks
| Поле | Значение |
|---|---|
| Baselines | NR |
| Retrospective | NR |
Prospective experimental validation
| Поле | Значение |
|---|---|
| Prospective | Experimental structural/functional evidence for general proteins |
| Tested designs | Multiple ligand/enzyme/interface campaigns; aggregate count: NR |
| Success criterion | Experimental structural/functional evidence for general proteins |
| Main results | Experimental structural/functional evidence for general proteins |
Limitations
Requires a credible backbone and ligand pose; does not generate backbone or prove catalysis/binding.
Что статья показывает
- Experimental structural/functional evidence for general proteins
Чего статья не показывает
- Predicted confidence, likelihood, pLDDT, docking or developability scores are not experimental affinity.
Размеченные утверждения
| Утверждение | Тип | Локатор |
|---|---|---|
| Atomic context-conditioned protein sequence design using LigandMPNN | Author-reported | DOI/URL from landscape matrix |
| Publication status: Peer-reviewed journal; Nature Methods (Q1) | Author-reported | landscape matrix, publication status column |
Source-grounded разбор статьи
Исследовательский вопрос
[MATRIX]Atomic-context inverse folding around small molecules, nucleotides and metals
Утверждения и поддержка evidence
[MATRIX]Claimed novelty: NR
[INTERPRETATION]Статья не «доказывает» общее преимущество метода вне своего экспериментального дизайна.
Что представленный evidence не поддерживает
- Predicted confidence, likelihood, pLDDT, docking or developability scores are not experimental affinity.
Ограничения, признанные авторами
[MATRIX]Requires a credible backbone and ligand pose; does not generate backbone or prove catalysis/binding.
Дополнительные ограничения интерпретации
NRПока Methods/SI не прочитаны, нельзя утверждать split leakage, точные гиперпараметры и полноту ablation.