Protein and Ligand Embeddings Generated by HuggingLigand (v1.0.0)

This dataset contains a small sample of protein and ligand embeddings generated using the HuggingLigand pipeline. It includes only 5 rows from the original processed dataset, selected for demonstration or testing purposes. The embeddings were created by running the following command:poetry run python src/scripts/cli.py --rows 5 Ligand embeddings were derived using seyonec/ChemBERTa-zinc-base-v1 on SMILES strings. Protein embeddings were generated using Rostlab/prot_t5_xl_half_uniref50-enc on amino acid sequences. These samples are based on the BindingDB dataset snapshot from July 2, 2025, archived at: https://zenodo.org/records/15800680 The data is provided for testing, demonstration, or reproducibility validation. It does not represent the full dataset used in model training.