ToxinPred2 is developed for predicting toxicity of proteins. In case user is interested in predicting toxicity of peptides then users should use our old server ToxinPred, which is specifically designed for peptides (See PLoS One. 2013 Sep 13;8(9):e73957). It utilizes wide range of information and techniques for prediction that includes machine learning techniques, BLAST, and MERCI. Our models have been trained on large dataset that contains 8233 toxic and 8233 non-toxic. In addition to following major modules, the server also has following important links; i) Download (Standalone Version & Datasets), ii) Algorithm, iii) Help page, iv) Team and v) Contact.