Datasets:
entry_id stringlengths 12 16 | uniprot_accession stringlengths 6 10 | tax_id int64 25 2.82M | organism_name stringlengths 7 143 | global_plddt float32 70 98.6 | seq_len int32 28 1.93k | seq_cluster_id stringlengths 6 10 | struct_cluster_id stringlengths 6 10 | split stringclasses 3
values | gcs_uri stringlengths 68 72 | cif_content stringlengths 36.9k 1.85M |
|---|---|---|---|---|---|---|---|---|---|---|
AF-A0A090SHW3-F1 | A0A090SHW3 | 990,271 | Vibrio variabilis | 88.25 | 89 | A0A090SHW3 | A0A1A9I5E1 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A090SHW3-F1-model_v4.cif | data_AF-A0A090SHW3-F1
#
_entry.id AF-A0A090SHW3-F1
#
loop_
_atom_type.symbol
C
N
O
S
#
loop_
_audit_author.name
_audit_author.pdbx_ordinal
"Jumper, John" 1
"Evans, Richard" 2
"Pritzel, Alexander" 3
"Green, Tim" 4
"Figurnov, Michael" 5
"Ronneberger... |
AF-A0A1C6JXL4-F1 | A0A1C6JXL4 | 765,821 | uncultured Blautia sp | 90.620003 | 270 | A0A1C6JXL4 | A0A1E7Z3J0 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A1C6JXL4-F1-model_v4.cif | "data_AF-A0A1C6JXL4-F1\n#\n_entry.id AF-A0A1C6JXL4-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
AF-A0A368QHG6-F1 | A0A368QHG6 | 4,555 | Setaria italica | 81.5 | 137 | A0A368QHG6 | A0A2J6LJ44 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A368QHG6-F1-model_v4.cif | "data_AF-A0A368QHG6-F1\n#\n_entry.id AF-A0A368QHG6-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
AF-A0A0F9MA88-F1 | A0A0F9MA88 | 412,755 | marine sediment metagenome | 80.5 | 66 | A0A0F9MA88 | A0A0F9MA88 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A0F9MA88-F1-model_v4.cif | "data_AF-A0A0F9MA88-F1\n#\n_entry.id AF-A0A0F9MA88-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
AF-A0A3D0YCS9-F1 | A0A3D0YCS9 | 2,049,046 | Porphyromonadaceae bacterium | 91.25 | 66 | A0A3D0YCS9 | A0A3D0YCS9 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A3D0YCS9-F1-model_v4.cif | "data_AF-A0A3D0YCS9-F1\n#\n_entry.id AF-A0A3D0YCS9-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
AF-A0A0D6M4T4-F1 | A0A0D6M4T4 | 53,326 | Ancylostoma ceylanicum | 92.440002 | 282 | A0A0D6M4T4 | A0A0J7P1G7 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A0D6M4T4-F1-model_v4.cif | "data_AF-A0A0D6M4T4-F1\n#\n_entry.id AF-A0A0D6M4T4-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
AF-K7EXM5-F1 | K7EXM5 | 13,735 | Pelodiscus sinensis | 88.75 | 293 | K7EXM5 | A0A8B6GPJ0 | train | gs://public-datasets-deepmind-alphafold-v4/AF-K7EXM5-F1-model_v4.cif | "data_AF-K7EXM5-F1\n#\n_entry.id AF-K7EXM5-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n#\nloop_(...TRUNCATED) |
AF-A0A7V9LZF8-F1 | A0A7V9LZF8 | 1,882,271 | Geodermatophilaceae bacterium | 84.25 | 195 | A0A7V9LZF8 | A0A7V9LZF8 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A7V9LZF8-F1-model_v4.cif | "data_AF-A0A7V9LZF8-F1\n#\n_entry.id AF-A0A7V9LZF8-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
AF-H0GCG4-F1 | H0GCG4 | 1,095,631 | Saccharomyces cerevisiae x Saccharomyces kudriavzevii (strain VIN7) | 71.25 | 241 | H0GCG4 | A0A2R4ACG4 | train | gs://public-datasets-deepmind-alphafold-v4/AF-H0GCG4-F1-model_v4.cif | "data_AF-H0GCG4-F1\n#\n_entry.id AF-H0GCG4-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n#\nloop_(...TRUNCATED) |
AF-A0A812XU89-F1 | A0A812XU89 | 230,985 | Symbiodinium sp. KB8 | 74.379997 | 467 | A0A812XU89 | A0A7S2E5A3 | train | gs://public-datasets-deepmind-alphafold-v4/AF-A0A812XU89-F1-model_v4.cif | "data_AF-A0A812XU89-F1\n#\n_entry.id AF-A0A812XU89-F1\n#\nloop_\n_atom_type.symbol\nC \nN \nO \nS \n(...TRUNCATED) |
End of preview. Expand in Data Studio
AFDB-1.6M — One Representative Structure Per Structural Cluster
A deduplicated subset of AFDB-24M, containing approximately 1.6 million AlphaFold Database predicted protein structures — one per structural cluster.
How This Dataset Was Created
This dataset was derived from AFDB-24M using the following procedure:
- All ~24 million rows across 12,005 shards were scanned.
- Rows were grouped by
struct_cluster_id(structural cluster representative from AFDB Foldseek clustering). - For each unique
struct_cluster_id, the single row with the highestglobal_plddt(global mean pLDDT confidence score) was selected. - The selected rows were written into new Parquet shards (2,000 rows each, ZSTD level 12 compression).
This yields approximately 1.6 million entries — one high-confidence representative per 3D structural fold cluster.
Dataset Summary
| Property | Value |
|---|---|
| Source | AFDB-24M |
| Total entries | ~1.6M (one per struct_cluster_id) |
| Selection criterion | Highest global_plddt per structural cluster |
| Format | Apache Parquet, ZSTD compressed (level 12) |
| Splits | train (98%), val (1%), test (1%) — inherited from AFDB-24M |
Schema
Each Parquet file contains a flat table with the following columns (same schema as AFDB-24M):
| Column | Type | Description |
|---|---|---|
entry_id |
string |
AFDB entry ID (e.g., AF-A0A1C0V126-F1) |
uniprot_accession |
string |
UniProt accession (e.g., A0A1C0V126) |
tax_id |
int64 |
NCBI taxonomy ID |
organism_name |
string |
Scientific name of the organism |
global_plddt |
float32 |
Global mean pLDDT confidence score (70–100) |
seq_len |
int32 |
Sequence length in residues |
seq_cluster_id |
string |
AFDB50 sequence cluster representative entry ID |
struct_cluster_id |
string |
Structural cluster representative entry ID |
split |
string |
train, val, or test |
gcs_uri |
string |
Original GCS URI |
cif_content |
string |
Complete raw mmCIF file text |
Usage
Loading with PyArrow
import pyarrow.parquet as pq
table = pq.read_table("shard_000000.parquet")
print(table.schema)
print(f"{len(table)} rows")
Loading with Pandas
import pandas as pd
df = pd.read_parquet("shard_000000.parquet")
print(df[["entry_id", "organism_name", "global_plddt", "seq_len", "split"]].head())
Parsing Structures with Gemmi
import gemmi
row = table.to_pydict()
cif_text = row["cif_content"][0]
doc = gemmi.cif.read_string(cif_text)
structure = gemmi.make_structure_from_block(doc.sole_block())
model = structure[0]
chain = model[0]
print(f"{len(chain)} residues")
Data Source and License
- AlphaFold Database structures are provided by DeepMind and EMBL-EBI under CC BY 4.0.
- Cluster files are from the Steinegger lab, based on Foldseek clustering of AFDB v4 (Version 3 clusters).
Citation
If you use this dataset, please cite the AlphaFold Database:
@article{varadi2022alphafold,
title={AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models},
author={Varadi, Mihaly and Anyango, Stephen and Deshpande, Mandar and others},
journal={Nucleic Acids Research},
volume={50},
number={D1},
pages={D439--D444},
year={2022},
doi={10.1093/nar/gkab1061}
}
And the AFDB cluster resource:
@article{barrio2024clustering,
title={Clustering predicted structures at the scale of the known protein universe},
author={Barrio-Hernandez, Inigo and Yeo, Jimin and Jänes, Jürgen and others},
journal={Nature},
volume={622},
pages={637--645},
year={2023},
doi={10.1038/s41586-023-06510-w}
}
- Downloads last month
- 779