HELI-Cas

High-dimensional Ensemble Learning for Identification of Cas proteins

A binary machine-learning classifier that predicts whether a single protein sequence is a CRISPR-associated (Cas) protein — using a multi-modal feature pipeline and a 5-fold stacked ensemble.

Binary: Cas vs Non-Cas

5-fold stacked ensemble

PSSM · HMM · physicochemical features

Export CSV · JSON · PDF

Start Analysis Documentation

Tool Workflow

Feature pipeline

PSSM · HMM · physico-chem · disorder

5-fold ensemble

CatBoost + LightGBM

Input FASTA

50–2000 aa

Prediction

Cas / Non-Cas + confidence

From one FASTA to a Cas / Non-Cas call.