Datasets:

Aurigene-AI
/

MoleculeNet_ESOL

license: unknown
task_categories:
  - tabular-regression
  - graph-ml
  - tabular-classification
tags:
  - chemistry
  - biology
  - medical
  - solubility
  - molecular-property-prediction
  - drug-discovery
  - aurigene
pretty_name: MoleculeNet ESOL
size_categories:
  - 1K<n<10K
configs:
  - config_name: default
    data_files:
      - split: train
        path: esol.csv

Mirrored by Aurigene AI

Discovery stage: Lead optimization

Aqueous solubility (log mol/L) for small organic molecules. Regression, the standard solubility benchmark.

Rows: 1,128 (esol.csv 1,128)

Pairs with Aurigene-AI/MoLFormer-XL-both-10pct from our model catalogue.

Upstream: scikit-fingerprints/MoleculeNet_ESOL - all credit to the original authors and to the researchers who produced the underlying data; the dataset card and licence below are theirs.

Explore the rest of the catalogue: Molecule Explorer - Protein Target Explorer - Drug Discovery Model Hub

MoleculeNet ESOL

ESOL (Estimated SOLubility) dataset [1], part of MoleculeNet [2] benchmark. It is intended to be used through scikit-fingerprints library.

The task is to predict aqueous solubility. Targets are log-transformed, and the unit is log mols per litre (log Mol/L).

Characteristic	Description
Tasks	1
Task type	regression
Total samples	1128
Recommended split	scaffold
Recommended metric	RMSE

References

[1] John S. Delaney "ESOL: Estimating Aqueous Solubility Directly from Molecular Structure" J. Chem. Inf. Comput. Sci. 2004, 44, 3, 1000–1005 https://pubs.acs.org/doi/10.1021/ci034243x

[2] Wu, Zhenqin, et al. "MoleculeNet: a benchmark for molecular machine learning." Chemical Science 9.2 (2018): 513-530 https://pubs.rsc.org/en/content/articlelanding/2018/sc/c7sc02664a