home/categories/computational-chemistry/lamm-mit-scienceclaw-skills-molfeat-skill-md
computational-chemistryresearch

molfeat

Molecular ML featurization library (100+ featurizers: ECFP, descriptors, ChemBERTa). Input: SMILES strings you already possess. Output: numerical feature vectors for QSAR/ML models. Does NOT retrieve compounds from any database — querying by topic name returns only a metadata stub. Use pubchem or chembl to obtain SMILES first, then featurize here. For ADMET predictions use tdc.

lamm-mit
maintainer
lamm-mit
更新日 3/14/2026
スター
156
フォーク
35
quick start

Installation and usage

Molecular ML featurization library (100+ featurizers: ECFP, descriptors, ChemBERTa). Input: SMILES strings you already possess. Output: numerical feature vectors for QSAR/ML models. Does NOT retrieve compounds from any database — querying by topic name returns only a metadata stub. Use pubchem or chembl to obtain SMILES first, then featurize here. For ADMET predictions use tdc.

インストール
$ install --globalskills.sh
使い方

インストール後、ターミナルで以下のコマンドを実行してこのスキルを使用できます:

skills use molfeat