Molfeat — molecular featurization hub
Turn molecules into ML-ready feature matrices with molfeat: ECFP/MACCS/MAP4 fingerprints, RDKit and Mordred descriptors, pharmacophore and shape descriptors, and pretrained embeddings (ChemBERTa, ChemGPT, CheMeleon) as scikit-learn transformers
- What
- Turn molecules into ML-ready feature matrices with molfeat: ECFP/MACCS/MAP4 fingerprints, RDKit and Mordred descriptors, pharmacophore and shape descriptors, and pretrained embeddings (ChemBERTa, ChemGPT, CheMeleon) as scikit-learn transformers
- Cost
- Free
- Needs
- Python >= 3.11 with uv — molfeat (uv pip install molfeat); optional extras per need: molfeat[transformer] (ChemBERTa/ChemGPT), molfeat[mordred], molfeat[fcd], molfeat[pyg]; separate map4 package from GitHub if using MAP4 fingerprints; your own SMILES data
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — @alterlab-ieu's alterlab-molfeat skill, listed here with credit to its creator (part of the AlterLab Academic Skills suite): a complete agent reference for molecular featurization with the molfeat Python library. It covers the three-layer model — calculators (per-molecule), transformers (scikit-learn compatible, batched, parallel), and pretrained transformers (ChemBERTa, ChemGPT, MolT5, CheMeleon, Mol-JEPA) — a featurizer selection guide by use case (traditional ML, deep learning, similarity search, pharmacophore approaches), common workflows (QSAR model building, virtual screening, scikit-learn pipeline integration, comparing featurizers), performance tips (parallelization, batching, caching, float32), state save/load for reproducibility, and error handling for invalid SMILES. Refreshingly honest about molfeat 1.0 breaking changes: DGL GIN/Graphormer and protein featurizers removed, no dgl extra, Mol-JEPA weights are CC BY-NC 4.0, and MAP4 needs the separate map4 package from GitHub. Honest caveats: license listed MIT per the discovery manifest (the skill's own frontmatter reads Apache-2.0 — the manifest makes faith here); pretrained foundation-model downloads can be large and slow on first run; your molecules come from you — the skill turns them into features, it trains nothing; outputs are research-grade features for your own models. Skill Harbor never reviews the code, review it yourself before use. Discovered via skills.sh.
Version:
Install
Prerequisites: Python >= 3.11 with uv — molfeat (uv pip install molfeat); optional extras per need: molfeat[transformer] (ChemBERTa/ChemGPT), molfeat[mordred], molfeat[fcd], molfeat[pyg]; separate map4 package from GitHub if using MAP4 fingerprints; your own SMILES data Install "Molfeat — molecular featurization hub" for me. It gives my agent @alterlab-ieu's molfeat reference: calculators, scikit-learn compatible transformers, and pretrained transformers for turning molecules into ML-ready feature vectors — featurizer selection guides (ECFP/MACCS/MAP4, RDKit/Mordred descriptors, pharmacophore/shape descriptors, ChemBERTa/ChemGPT/CheMeleon/Mol-JEPA embeddings), quick-start workflows (QSAR, virtual screening, similarity search, sklearn pipeline integration), performance tips, reproducibility via state YAML, and error handling — with honest notes on the molfeat 1.0 breaking changes. Part of the AlterLab Academic Skills suite. Listed MIT per the discovery manifest (the skill's frontmatter reads Apache-2.0 — the manifest makes faith). Repository: https://github.com/alterlab-ieu/alterlab-academic-skills/blob/main/skills/cheminformatics/alterlab-molfeat/SKILL.md 1. Fetch the SKILL.md file (and any helper files) from the repository path into a temporary folder and summarize what it does in one or two sentences. 2. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). This repo should contain zero secrets in code. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "alterlab-molfeat". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. uv pip install molfeat plus the extras I need; provide my own SMILES data; save featurizer configs to YAML for reproducibility). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.