Description
molfeat is an agent skill for featurising small molecules with Molfeat, for QSAR and QSPR models, chemical similarity, virtual screening and molecular machine learning. It turns SMILES strings or RDKit molecules into fingerprints, descriptors, pharmacophores or pretrained embeddings. It covers fingerprints such as ECFP and MACCS, and RDKit descriptors among others. The skill advises comparing representations on the same molecular split and the same assay endpoint, so that the comparison is fair. It targets Molfeat 1.0.0. The skill warns that the earlier 0.11 guidance is obsolete. Version 1.x supports modern Python and removes DGL and DGLLife, the legacy Graphormer models and the protein adapters. Some older model-store cards can still name adapters that have been removed, and the tagged 1.0 migration guide and source take precedence over older documentation still served at the stable address. For installation, the skill recommends an isolated environment. Its core examples were run on Python 3.13.3 on macOS with Apple Silicon, using Molfeat 1.0.0, datamol 0.13.0, RDKit 2026.03.6 and PyTorch 2.14.1.…
Highlights
- Fingerprints, descriptors and pretrained embeddings
- ECFP and MACCS fingerprints
- Fair comparison on the same split
- Targets Molfeat 1.0.0
- Migration notes from the 0.11 guidance
Ratings & Reviews
No ratings yet
Be the first to rate molfeat.
People also visit
pytorch-patterns
PyTorch deep learning patterns and best practices for building robust
recsys-pipeline-architect
Design composable recommendation, ranking, and feed pipelines
datamol
Pythonic wrapper around RDKit with simplified interface and sensible defaults
molfeat FAQ
What does molfeat do?
It turns SMILES or RDKit molecules into fingerprints, descriptors, pharmacophores or pretrained embeddings for machine learning and similarity search.
Which fingerprints are covered?
ECFP and MACCS, among others described in the skill.
What changed in version 1?
Version 1.x supports modern Python and removes DGL and DGLLife, legacy Graphormer and the protein adapters.
Which environment was used for the examples?
Python 3.13.3 on macOS Apple Silicon, with Molfeat 1.0.0, datamol 0.13.0, RDKit 2026.03.6 and PyTorch 2.14.1.
How do I install it?
Run npx skills add https://github.com/k-dense-ai/scientific-agent-skills --skill molfeat.
Is it checked for security?
Its security audits show passes from Gen Agent Trust Hub, Socket and Snyk.
What are the best molfeat alternatives?
Popular alternatives to molfeat include ml-adoption-playbook, mle-workflow and pytorch-patterns. See all molfeat alternatives