deepchem

Molecular ML with diverse featurizers and pre-built datasets. Use for property prediction (ADMET, toxicity) with traditional ML or GNNs when you want extensive featurization options and MoleculeNet benchmarks. Best for quick experiments with pre-trained models, diverse molecular representations. For graph-first PyTorch workflows use torchdrug; for benchmark datasets use pytdc.

1.36x

Quality

75%

Does it follow best practices?

Impact

83%

1.36x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./scientific-skills/deepchem/SKILL.md

Evaluation results

84%

Molecular Solubility Prediction Pipeline

Small-dataset featurizer and model selection

Criteria

Without context

With context

Scaffold splitter used

100%

CircularFingerprint radius

100%

CircularFingerprint size

100%

Simple model chosen

100%

SklearnModel wrapper

100%

NormalizationTransformer applied

80%

transform_y=True set

100%

Transformers applied to new molecules

66%

50%

CSVLoader feature_field

100%

Three-way split

NumpyDataset for new molecules

100%

79%

12%

Multi-Task Molecular Toxicity Classifier

Medium-dataset toxicity classification with imbalanced data

Criteria

Without context

With context

Scaffold splitter selected

100%

No GNN for medium dataset

100%

Medium-dataset model

50%

Fingerprint featurizer for non-GNN

100%

Imbalance handling

50%

100%

MoleculeNet Tox21 loader

100%

ROC-AUC metric

50%

Multitask evaluation

100%

CircularFingerprint params

100%

50%

evaluation_results.json present

100%

86%

47%

Blood-Brain Barrier Permeability Prediction with Pretrained Models

Transfer learning on small novel-scaffold molecular dataset

Criteria

Without context

With context

HuggingFaceModel used

100%

Correct ChemBERTa model ID

71%

Low fine-tuning learning rate

85%

100%

DummyFeaturizer for pretrained model

100%

Scaffold splitter used

50%

100%

CSVLoader with feature_field

100%

Approach summary written

100%

uv installation command

n_tasks set to 1

Classification task type

50%

100%

Repository: K-Dense-AI/claude-scientific-skills
Commit: b58ad7e

Evaluated: 27 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Molecular Solubility Prediction Pipeline Multi-Task Molecular Toxicity Classifier Blood-Brain Barrier Permeability Prediction with Pretrained Models

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.