Skills deepchem
📦

deepchem

Content revision r1 Safe ⚙️ External commands🌐 Network access

Build Molecular ML Workflows with DeepChem

Molecular machine learning often requires specialized featurizers, datasets, and evaluation methods. This skill helps Claude, Codex, and Claude Code design DeepChem workflows for property prediction, transfer learning, and benchmarks.

Supports: Claude Codex Code(CC)
🥉 79 Bronze

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request
Review the Skillstore skill "deepchem" from https://skillstore.io/skills/k-dense-ai-deepchem.md and its manifest at https://skillstore.io/api/skills/k-dense-ai-deepchem/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/k-dense-ai-deepchem.md Signed manifest GET /api/skills/k-dense-ai-deepchem/manifest Signed lockfile GET /api/skills/k-dense-ai-deepchem/lockfile

Test it

Using "deepchem". I have 800 molecules with SMILES and solubility labels. What should I try first?

Expected outcome:

  • Start with CircularFingerprint and a simple regression model as a baseline.
  • Use ScaffoldSplitter to reduce chemical leakage between train and test sets.
  • Consider transfer learning if the baseline is unstable or the labels are sparse.

Using "deepchem". I need to evaluate toxicity prediction on Tox21.

Expected outcome:

  • Use a MoleculeNet Tox21 loader with graph features and scaffold splitting.
  • Train a multitask classifier because Tox21 has multiple toxicity tasks.
  • Report ROC-AUC per task and include missing-label handling in the evaluation notes.

Using "deepchem". My graph model is running out of memory.

Expected outcome:

  • Reduce batch size before changing the model architecture.
  • Use disk-backed datasets for large feature arrays.
  • Start with a fingerprint baseline to confirm the dataset and labels are usable.

Security Audit

Safe
v9 • 7/9/2026 Open versioned report

All static findings were adjudicated as false positives after reviewing the cited files. The alerts come from Markdown code fences, DeepChem API names, ML validation terminology, argparse choices, and public documentation links rather than malicious behavior.

7
Files scanned
2,524
Lines analyzed
0
Review items
0
False positives ignored

Risk Factors

⚙️ External commands (50)
SKILL.md:34-53 SKILL.md:53-56 SKILL.md:56-57 SKILL.md:57-58 SKILL.md:58-59 SKILL.md:59-60 SKILL.md:60-68 SKILL.md:68-90 SKILL.md:90-94 SKILL.md:94-106 SKILL.md:106-114 SKILL.md:114-118 SKILL.md:118-120 SKILL.md:120-137 SKILL.md:137-140 SKILL.md:140-141 SKILL.md:141-142 SKILL.md:142-143 SKILL.md:143-144 SKILL.md:144-160 SKILL.md:160-167 SKILL.md:167-170 SKILL.md:170-180 SKILL.md:180-183 SKILL.md:183-200 SKILL.md:200-206 SKILL.md:206-221 SKILL.md:221-224 SKILL.md:224 SKILL.md:225 SKILL.md:226 SKILL.md:227 SKILL.md:229-235 SKILL.md:235-251 SKILL.md:251-259 SKILL.md:259-263 SKILL.md:263-280 SKILL.md:280-284 SKILL.md:284-298 SKILL.md:298-306 SKILL.md:306-324 SKILL.md:324-330 SKILL.md:330-367 SKILL.md:367-373 SKILL.md:373-401 SKILL.md:401-403 SKILL.md:403-407 SKILL.md:407-409 SKILL.md:409-412 SKILL.md:412-422
🌐 Network access (3)
SKILL.md:591 SKILL.md:592 SKILL.md:593
No confirmed security findings were detected by the latest completed static and semantic audit. This does not prove the skill has no side effects.
Audited by: claude View Audit History →
Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report
Security Assessment

Copy report link

https://skillstore.io/skills/k-dense-ai-deepchem/audits/9?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/k-dense-ai-deepchem/security.svg)](https://skillstore.io/skills/k-dense-ai-deepchem?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/k-dense-ai-deepchem?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/k-dense-ai-deepchem/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/k-dense-ai-deepchem.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Academic citations (APA · BibTeX · CFF)

APA citation

K-Dense-AI. (2026). deepchem security audit report (audit version 9) [Author version unspecified]. Skillstore. https://skillstore.io/skills/k-dense-ai-deepchem/audits/9

BibTeX citation

@techreport{k-dense-ai-k-dense-ai-deepchem-2026, author = {K-Dense-AI}, title = {deepchem security audit report (audit version 9)}, institution = {Skillstore}, year = {2026}, number = {9}, url = {https://skillstore.io/skills/k-dense-ai-deepchem/audits/9}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "deepchem security audit report (audit version 9)" version: "unspecified" type: report authors: - name: "K-Dense-AI" date-released: "2026-07-09" url: "https://skillstore.io/skills/k-dense-ai-deepchem/audits/9" identifiers: - type: other value: "skillstore:k-dense-ai-deepchem:audit:9" description: "Skillstore immutable audit report identifier"

Compare variants

2 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Higher Skillstore usage
K-Dense-AI Recommended Current

k-dense-ai-deepchem

Skillstore Score 79
Evidence Confidence High
Skillstore usage 10
Updated

2026-08-21

davila7

davila7-deepchem

Skillstore Score 79
Evidence Confidence High
Skillstore usage 7
Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: High
64
Architecture
85
Maintainability
87
Content
70
Community
78
Spec Compliance

What You Can Build

Prioritize drug candidates

Estimate solubility, toxicity, or activity trends before costly laboratory screening.

Prototype molecular models

Compare fingerprints, descriptors, and graph neural networks on a custom molecular dataset.

Teach chemistry ML workflows

Create guided examples that explain featurization, scaffold splitting, metrics, and benchmark evaluation.

Try These Prompts

Choose a featurizer 
Help me choose a DeepChem featurizer for a CSV file with SMILES strings and one solubility target. Explain the tradeoffs.
Train a benchmark model 
Design a DeepChem workflow to train a graph convolution model on Tox21 with scaffold splitting and ROC-AUC evaluation.
Use transfer learning 
Plan a transfer learning workflow for a small binding affinity dataset using a pretrained molecular model and proper validation.
Compare model strategies 
Compare CircularFingerprint, RDKitDescriptors, GCN, and AttentiveFP for my ADMET dataset. Recommend an experiment plan and metrics.

Best Practices

  • Use scaffold splitting for molecular prediction tasks to reduce structure-based data leakage.
  • Start with simple fingerprint baselines before training deep graph neural networks.
  • Match metrics to the task type, such as ROC-AUC for classification and RMSE or R2 for regression.

Avoid

  • Do not use random splits for drug discovery claims when similar scaffolds may appear in both sets.
  • Do not train large graph neural networks on very small datasets without strong regularization or transfer learning.
  • Do not treat predicted molecular properties as experimental evidence without external validation.

Frequently Asked Questions

What is this skill best for?
It is best for planning and explaining DeepChem workflows for molecular property prediction, benchmarks, and transfer learning.
Can it run DeepChem models directly?
It can guide code and scripts, but execution requires a local environment with DeepChem and its dependencies installed.
Which featurizer should I start with?
Start with CircularFingerprint for a fast baseline, then try graph featurizers when you have enough data.
Why is scaffold splitting recommended?
Scaffold splitting keeps related molecular structures together and gives a more realistic estimate of generalization.
Does it support pretrained models?
Yes. The documentation covers transfer learning patterns for models such as ChemBERTa, GROVER, and MolFormer.
Is this a substitute for laboratory validation?
No. The skill helps build computational workflows, but results should be validated with appropriate scientific methods.

Developer Details

Author

K-Dense-AI

License

MIT license

Skillstore revision

r1

Version notice

The author did not declare a version.

Repository

https://github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/deepchem

Ref

1ffa7643651792ccb4bd3b15d924d2c97edff755

Maintenance freshness

7/18/2026

Usage

9 downloads · 202 views

File structure

📁 references/

📄 api_reference.md

📄 workflows.md

📁 scripts/

📄 graph_neural_network.py

📄 predict_solubility.py

📄 transfer_learning.py

📄 evaluation.json

📄 SKILL.md