CHEESESearch API

On-Premise Endpoints

Embeddings, centroids, pairwise/matrix similarity, and visualization endpoints available exclusively with an on-premise CHEESE installation.

Suggest an edit

The following endpoints are available exclusively with an on-premise CHEESE installation. Replace YOUR_ONPREM_URL with your deployment's base URL.

GET /embeddings — Compute Embeddings

Compute CHEESE embeddings (256-dimensional vectors) for molecules.

params = {
    "search_input": ["Fc1ccccc1", "Clc1ccccc1"],
    "save_embs": False,
}

response = requests.get(MY_URL + "/embeddings",
    params=params, headers=headers)
# Returns 256-dim vectors per molecule per search type

GET /centroid_embeddings — Centroid Embeddings

Retrieve cluster centroid embeddings for databases. Optionally include representative molecules for each cluster.

params = {
    "db_name": "ZINC15",
    "search_type": "espsim_shape",
    "centroid_mols": False,
}

response = requests.get(MY_URL + "/centroid_embeddings",
    params=params, headers=headers)

GET /similarity — Pairwise Similarity

Compute pairwise similarity between two molecules using euclidean or cosine distance across all similarity metrics.

params = {
    "smiles1": "Fc1ccccc1",
    "smiles2": "Clc1ccccc1",
    "similarity_metric": "all",
    "distance_type": "cosine",
}

response = requests.get(MY_URL + "/similarity",
    params=params, headers=headers)

GET /similarity_matrix — Similarity Matrix

Compute a full pairwise similarity matrix for a list of molecules. Useful for clustering and network analysis.

params = {
    "smiles": ["SMILES1", "SMILES2", "SMILES3"],
    "similarity_metric": "espsim_shape",
    "distance_type": "cosine",
}

response = requests.get(MY_URL + "/similarity_matrix",
    params=params, headers=headers)

GET /visualise — Visualization

Generate 2D coordinates for visualization using UMAP or PCA dimensionality reduction. Supports saving coordinates to disk.

params = {
    "search_input": ["Fc1ccccc1", "Clc1ccccc1", "Brc1ccccc1"],
    "search_type": "espsim_electrostatic",
    "visualisation_method": "umap",
}

response = requests.get(MY_URL + "/visualise",
    params=params, headers=headers)

On this page