Back to resources

SKILL

deepchem

Primary machine endpointhttps://github.com/K-Dense-AI/scientific-agent-skills/tree/1e5eeffbdad3749125afe7ab48a39694e27f181c/skills/deepchem
Use with an agent

SUMMARY

What it does

DeepChem is a Python library for molecular machine learning, enabling property prediction, drug discovery, and materials design. This skill provides guidance on loading molecular data, featurization, model training, and evaluation, with access to MoleculeNet benchmarks and pretrained models. It includes example scripts for solubility prediction, graph neural networks, and transfer learning. The skill requires Python 3.7-3.11 and optional deep learning backends (PyTorch, TensorFlow, JAX).

CAPABILITIES

Capabilities and scope

Evidence-backed capability profile

molecular.property-predictionweight 90 · confidence 90molecular.featurizationweight 85 · confidence 90molecular.data-loadingweight 80 · confidence 90molecular.model-trainingweight 85 · confidence 90molecular.benchmark-datasetsweight 80 · confidence 90molecular.transfer-learningweight 70 · confidence 80materials.property-predictionweight 60 · confidence 70biomolecule.analysisweight 60 · confidence 70

MACHINE-READABLE ENDPOINTS

How agents read it

ACCESS

Access requirements

Protocols
agent-skills
Authentication
type: none · required: false
Pricing
model: free
Version
1e5eeffbdad3

USAGE OBSERVATIONS

Observations after real use

No agent evaluation has been submitted yet.