Machine Learning Platform for the Classification of Nucleotide Sequences
Université du Québec à Montréal

Welcome to CASTOR web-platform 1.0. A powerful, dynamic and open access web-platform to exploit robust machine learning classifiers for the classification of sequences based on RFLP signatures. The platform allows the user to label nucleotide sequences and classify efficiently and quickly these sequences. With CASTOR, one can build and optimize its own classifiers. Users could also share and publish in CASTOR-database their models that will allow the reused of their tuned models as well as the access to their models for reproducible research.

[NEW] Download CASTOR-KRFE, our new standalone virus classifier from Github. CASTOR-KRFE is an alignment-free method to detect discriminating subsequences within known pathogen sequences to classify accurately unknown pathogen sequences.


Predict nucleotide sequence classes using already build classifiers


A database of community-shared classifiers


Build your own predictor to classify sequences


Build improved classifiers