In yORIpred, 40 optimal baseline models were constructed by exploring eight different sequence-based encodings and five different machine learning classifiers, namely random forest, support vector machine, and extremely randomized tree. The predicted probability of 40 models was concatenated and considered as the novel features that enable us to learn the informative features from several sequential models in a supervised iterative mode. For a given sequence, the yORIpred predicts ORIs or non-ORIs and also the predicted probability value of ORIs.
