{
  "READY": true,
  "name": "proteinea_solubility",
  "task": "binary",
  "num_classes": 2,
  "cat_idx": [],
  "num_samples": 71421,
  "num_features": 8000,
  "imbalance_ratio": 1.3829,
  "description": "Sample-level binary classification task predicting protein solubility. Labels: 1 = soluble, 0 = insoluble. Features are k-mer (k=3) frequency vectors computed from single amino-acid sequences. Missing feature values are not imputed.",
  "source": "proteinea/solubility (Hugging Face). Binary protein solubility prediction dataset. Labels indicate whether a protein is soluble (1) or insoluble (0). Separate train, validation and test CSV files.",
  "label": "Proteomics",
  "sub_labels": [
    "protein-seq"
  ]
}