Datasets:

LoocasGoose
/

cpr_data

license: apache-2.0
task_categories:
  - feature-extraction
  - text-classification
tags:
  - protein
  - bioinformatics
  - embeddings
  - conformal-prediction
size_categories:
  - 1M<n<10M

Conformal Protein Retrieval - Data Files

This dataset contains the large data files required to run the Conformal Protein Retrieval Gradio Space.

📊 Lookup Databases

UniProt Database:

data/lookup_embeddings.npy - Pre-embedded UniProt protein sequences (Protein-Vec embeddings)
data/lookup_embeddings_meta_data.tsv - Metadata for UniProt proteins (Entry, Pfam, Protein names)

SCOPE Database:

data/lookup/scope_lookup_embeddings.npy - Pre-embedded SCOPE protein domain sequences
data/lookup/scope_lookup.fasta - FASTA metadata for SCOPE proteins

🎯 Conformal Prediction Files

results/fdr_thresholds.csv - Precomputed FDR (False Discovery Rate) thresholds
results/fnr_thresholds.csv - Precomputed FNR (False Negative Rate) thresholds
results/calibration_probs.csv - Calibration probabilities for Venn-Abers prediction

🧬 Protein-Vec Model

protein_vec_models/protein_vec.ckpt - Main Protein-Vec model checkpoint
protein_vec_models/protein_vec_params.json - Model configuration
protein_vec_models/*.py - Model architecture code files

Usage

These files are automatically loaded by the Gradio Space application. To use them locally:

from huggingface_hub import hf_hub_download
import numpy as np

# Download a specific file
embedding_file = hf_hub_download(
    repo_id="LoocasGoose/cpr_data",
    filename="data/lookup_embeddings.npy",
    repo_type="dataset"
)

# Load the embeddings
embeddings = np.load(embedding_file)

Citation

If you use these data files, please cite the original paper:

@article{boger2025functional,
  title={Functional protein mining with conformal guarantees},
  author={Boger, Ron S and Chithrananda, Seyone and Angelopoulos, Anastasios N and Yoon, Peter H and Jordan, Michael I and Doudna, Jennifer A},
  journal={Nature Communications},
  volume={16},
  number={1},
  pages={85},
  year={2025},
  publisher={Nature Publishing Group UK London}
}

License

Apache 2.0

Source

Original data from: Zenodo