Hugging Face's logo

Datasets:
Anonymoususer2223
/
ProtEnv

Tags:
biology
protein
protein-structure
License:
Dataset card Files
xet
 Community
1
Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 80, in _split_generators
                  raise ValueError(
                  ...<2 lines>...
                  )
              ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

ProtEnv: predicted structures for four ProtCompass downstream tasks

ESMFold structures for the four downstream tasks of ProtCompass whose source datasets provide no structure: stability, fluorescence, solubility and PPI affinity. The structure encoders read these files on these four tasks.

Contents

File Task PDB files Size (GB) Extracts to
predicted_structures/predictions_train.tar.gz stability, training split 52,029 0.36 predictions_train/
predicted_structures/predictions_test.tar.gz stability, test split 12,851 0.09 predictions_test/
predicted_structures/fluorescence.tar.gz fluorescence 54,025 2.01 predictions_fluorescence/
predicted_structures/solubility.tar.gz solubility 70,201 2.70 predictions_solubility/
predicted_structures/ppi_affinity.tar.gz PPI affinity, one structure per chain 2,601 0.05 predictions_ppi_affinity/

Each archive holds one folder of PDB files. A file holds one predicted chain as ATOM records, with the per-residue pLDDT (0 to 100) in the B-factor column.

Use with the code

The archives hold every structure that was predicted. The task manifests in the embeddings repository (extract/manifest/<task>.csv) select the proteins each task uses and give each protein's structure file in the column structure_path, relative to the directory that holds the code repository's protcompass/ folder. Extract each archive into the folder its manifest names:

Task Folder named in structure_path
stability downstream_tasks/data/tape/tape_structures/stability/predictions_train/ and .../predictions_test/
fluorescence protcompass/extract/structures/predictions_fluorescence/
solubility protcompass/extract/structures/predictions_solubility/
PPI affinity protcompass/extract/structures/predictions_ppi_affinity/ 
from huggingface_hub import hf_hub_download
import tarfile

path = hf_hub_download("Anonymoususer2223/ProtEnv", "predicted_structures/fluorescence.tar.gz", repo_type="dataset")
tarfile.open(path).extractall("protcompass/extract/structures/")

Notes

  • The structures are predictions, not experimental structures. The manifests record each protein's mean pLDDT (mean_plddt). Its median over the manifest rows is 70.0 for stability, 46.4 for fluorescence, 83.7 for solubility and 92.4 for PPI affinity.
  • The experimental (PDB), AlphaFold DB, SCOPe and CATH structures of the other tasks are public and are not redistributed here. The code repository has the scripts that fetch them.

Citation

Citation information will be added after the anonymous review period.
Downloads last month
27
Total file size:
5.22 GB

Collection including Anonymoususer2223/ProtEnv