Aqueous Solubility Database (AqSolDB)

AqSolDB is created by the Autonomous Energy Materials Discovery [AMD] research group, consists of aqueous solubility values of 9,982 unique compounds curated from 9 different publicly available aqueous solubility datasets. This openly accessible dataset, which is the largest of its kind, and will not only serve as a useful reference source of measured solubility data, but also as a much improved and generalizable training data source for building data-driven models.

Quickstart Usage

Load a dataset in python

Each subset can be loaded into python using the Huggingface datasets library. First, from the command line install the datasets library

$ pip install datasets

then, from within python load the datasets library

>>> import datasets

and load one of the AqSolDB datasets, e.g.,

>>> AqSolDB = datasets.load_dataset("maomlab/AqSolDB", name = "AqSolDB")
Downloading readme: 100%|████████████████████| 10.2k/10.2k [00:00<00:00, 4.41MB/s]
Downloading data: 100%|█████████████████████████| 972k/972k [00:02<00:00, 432kB/s]
Downloading data: 100%|██████████████████████| 2.88M/2.88M [00:01<00:00, 1.92MB/s]
Generating test split: 100%|████████| 2494/2494 [00:00<00:00, 44727.48 examples/s]
Generating train split: 100%|██████| 7488/7488 [00:00<00:00, 144316.82 examples/s]

and inspecting the loaded dataset

>>> AqSolDB
AqSolDB
DatasetDict({
    test: Dataset({
        features: ['ID', 'Name', 'InChI', 'InChIKey', 'SMILES', 'Y', 'SD', 'Ocurrences', 'Group', 'MolWt', 'MolLogP', 'MolMR', 'HeavyAtomCount', 'NumHAcceptors', 'NumHDonors', 'NumHeteroatoms', 'NumRotatableBonds', 'NumValenceEl\
ectrons', 'NumAromaticRings', 'NumSaturatedRings', 'NumAliphaticRings', 'RingCount', 'TPSA', 'LabuteASA', 'BalabanJ', 'BertzCT', 'ClusterNo', 'MolCount', 'group'],
        num_rows: 2494
    })
    train: Dataset({
        features: ['ID', 'Name', 'InChI', 'InChIKey', 'SMILES', 'Y', 'SD', 'Ocurrences', 'Group', 'MolWt', 'MolLogP', 'MolMR', 'HeavyAtomCount', 'NumHAcceptors', 'NumHDonors', 'NumHeteroatoms', 'NumRotatableBonds', 'NumValenceEl\
ectrons', 'NumAromaticRings', 'NumSaturatedRings', 'NumAliphaticRings', 'RingCount', 'TPSA', 'LabuteASA', 'BalabanJ', 'BertzCT', 'ClusterNo', 'MolCount', 'group'],
        num_rows: 7488
    })
})

Use a dataset to train a model

One way to use the dataset is through the MolFlux package developed by Exscientia. First, from the command line, install MolFlux library with catboost and rdkit support

pip install 'molflux[catboost,rdkit]'

then load, featurize, split, fit, and evaluate the catboost model

import json
from datasets import load_dataset
from molflux.datasets import featurise_dataset
from molflux.features import load_from_dicts as load_representations_from_dicts
from molflux.splits import load_from_dict as load_split_from_dict
from molflux.modelzoo import load_from_dict as load_model_from_dict
from molflux.metrics import load_suite

split_dataset = load_dataset('maomlab/AqSolDB')

split_featurised_dataset = featurise_dataset(
  split_dataset,
  column = "SMILES",
  representations = load_representations_from_dicts([{"name": "morgan"}, {"name": "maccs_rdkit"}]))

model = load_model_from_dict({
    "name": "cat_boost_regressor",
    "config": {
        "x_features": ['SMILES::morgan', 'SMILES::maccs_rdkit'],
        "y_features": ['Y']}})

model.train(split_featurised_dataset["train"])
preds = model.predict(split_featurised_dataset["test"])

regression_suite = load_suite("regression")

scores = regression_suite.compute(
    references=split_featurised_dataset["test"]['Y'],
    predictions=preds["cat_boost_regressor::Y"])    

Aqueous Solubility Database

Data splits

The original AqSoDB dataset does not define splits, so here we have used the Realistic Split method described in (Martin et al., 2018).

Citation

TY  - JOUR AU  - Sorkun, Murat Cihan AU  - Khetan, Abhishek AU  - Er, S√ºleyman PY  - 2019 DA  - 2019/08/08 TI  - AqSolDB, a curated reference set of aqueous solubility and 2D descriptors for a diverse set of compounds JO  - Scientific Data SP  - 143 VL  - 6 IS  - 1 AB  - Water is a ubiquitous solvent in chemistry and life. It is therefore no surprise that the aqueous solubility of compounds has a key role in various domains, including but not limited to drug discovery, paint, coating, and battery materials design. Measurement and prediction of aqueous solubility is a complex and prevailing challenge in chemistry. For the latter, different data-driven prediction models have recently been developed to augment the physics-based modeling approaches. To construct accurate data-driven estimation models, it is essential that the underlying experimental calibration data used by these models is of high fidelity and quality. Existing solubility datasets show variance in the chemical space of compounds covered, measurement methods, experimental conditions, but also in the non-standard representations, size, and accessibility of data. To address this problem, we generated a new database of compounds, AqSolDB, by merging a total of nine different aqueous solubility datasets, curating the merged data, standardizing and validating the compound representation formats, marking with reliability labels, and providing 2D descriptors of compounds as a Supplementary Resource. SN  - 2052-4463 UR  - https://doi.org/10.1038/s41597-019-0151-1 DO  - 10.1038/s41597-019-0151-1 ID  - Sorkun2019 ER  -