Prediction of Toxins

This page allows users to submit multiple sequences in FASTA format. The server will then predict whether the peptides are toxic or non-toxic based on their primary sequence. This page implements machine learning models, such as Extra tree (ET), deep learning neural network model includes ANN with LSTM, and their hybrid approaches (ET+MERCI and DL+MERCI). The user can select one of the models for prediction. By default, the server will use the ML+MERCI hybrid or ensemble approach, which has been shown to be the most accurate in our evaluations. For more information, please see the help page.