File size: 2,952 Bytes
9a31c5d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 | ---
license: cc0-1.0
tags:
- pdb
- protein
language:
- en
- zh
---
<p align="center">
<strong>
<span style="font-size: 30px;">ProteinMPNN Dataset</span>
</strong>
</p>
## Dataset Description
The ProteinMPNN training sample dataset contains PDB-derived structural data for protein sequence design tasks. The data directory is `pdb_2021aug02_sample`, which contains a structure index table, validation and test cluster split files, and `.pt` structure tensor files organized by PDB and chain.
## Supported Tasks
This resource is intended for validating training data loading, debugging the training workflow, and performing minimal data sanity checks for the `OneScience/ProteinMPNN` model. Model inference generally does not require this training dataset.
## Dataset Format and Structure
| Path | Description |
|---|---|
| `pdb_2021aug02_sample/list.csv` | Chain index with the header `CHAINID,DEPOSITION,RESOLUTION,HASH,CLUSTER,SEQUENCE` |
| `pdb_2021aug02_sample/valid_clusters.txt` | Validation cluster list |
| `pdb_2021aug02_sample/test_clusters.txt` | Test cluster list |
| `pdb_2021aug02_sample/pdb/` | PDB-derived `.pt` structure tensor files; the current sample contains 870 `.pt` files |
Each chain file typically contains fields such as the amino acid sequence, atomic coordinates, mask, temperature factors, and occupancy.
## How to Use the Dataset
### Files and Download
```bash
hf download --dataset OneScience-Sugon/proteinmpnn --local-dir ./proteinmpnn_dataset
```
Training and debugging require the model repository to be prepared separately:
```bash
hf download --model OneScience-Sugon/ProteinMPNN
```
### Data Placement and Validation
It is recommended to use the download directory as the data root:
```bash
export DATA_ROOT=/path/to/proteinmpnn_dataset
python tools/validate_dataset.py --data-root .
```
Example ProteinMPNN training script command:
```bash
python train/training.py --path_for_training_data "$DATA_ROOT/pdb_2021aug02_sample" --debug true
```
The training script is part of the compatible model repository and is not included in this dataset repository.
## Official OneScience Information
| Platform | OneScience Main Repository | Skills Repository |
|---|---|---|
| Gitee | https://gitee.com/onescience-ai/onescience | https://gitee.com/onescience-ai/oneskills |
| GitHub | https://github.com/onescience-ai/OneScience | https://github.com/onescience-ai/oneskills |
## Limitations and License
This dataset was curated and converted from wwPDB/PDB biological assemblies and is licensed under CC0-1.0.
## Limitations and License
This resource is a sample-scale dataset and does not represent the complete PDB training set. It is intended primarily for ProteinMPNN data loading, training debugging, and workflow sanity checks. When using it, cite the relevant ProteinMPNN papers, upstream projects, and original PDB data sources, and comply with the ModelScope platform rules.
|