Drug Discovery Model Hub

Every open model in the Aurigene AI catalogue, mapped onto the stage of drug discovery where it actually earns its keep. Click a stage to filter, click a card to see a copy-paste snippet. Download counts and likes are pulled live from the Hugging Face Hub.

Fetching live Hub statistics…

Interactive tools

How to use these models

⬇️

Load any model

transformers + huggingface_hub

pip install transformers torch from transformers import AutoModel, AutoTokenizer m = "Aurigene-AI/MoLFormer-XL-both-10pct" tok = AutoTokenizer.from_pretrained(m, trust_remote_code=True) mdl = AutoModel.from_pretrained(m, trust_remote_code=True)

🧪

Embed a molecule library

SMILES → vectors for QSAR / clustering

smiles = ["CC(=O)Oc1ccccc1C(=O)O", "CN1CCN(CC1)c1ccccc1"] batch = tok(smiles, padding=True, return_tensors="pt") emb = mdl(**batch).pooler_output # (2, 768)

🧬

Score a protein sequence

ESM-2 embeddings per residue

from transformers import AutoTokenizer, AutoModel m = "Aurigene-AI/esm2_t33_650M_UR50D" tok = AutoTokenizer.from_pretrained(m) esm = AutoModel.from_pretrained(m) out = esm(**tok("MQIFVKTLTGKTITLEVEPS", return_tensors="pt")) print(out.last_hidden_state.shape) # (1, L, 1280)