deepchem
Molecular predictions with meaningful holdouts.
What it does
Develops and evaluates molecular-property prediction workflows with DeepChem using explicit data, model and holdout-validation contracts.
When to use it
Use for molecular-property prediction, molecular featurization, MoleculeNet benchmarks and explicitly configured encoder transfer.
How to use it
Give it a molecular-property prediction task; it trains and evaluates a DeepChem model and reports metrics and predictions.
Uses
Good to know
- advertises something the task does not need
What you need · 3
Requires Python 3.11 for the tested DeepChem 2.8.0 stack, using an isolated environment when repository Python requirements conflict.
The workflow targets DeepChem 2.8.0 and warns against mixing development documentation with the stable installation.
Access · 5
This skill
DeepChem
Installs the tested DeepChem version
Reads DeepChem installation documentation
arXiv
Fetches the skill library paper