A number of computational tools have been developed to predict the presence of SPs. These programs have been greatly refined in recent years and their accuracy has improved with the adoption of machine learning algorithms. However, while this is a valuable approach, as with all prediction-based tools, there is a consistent error rate and both false positive and negative predictions should be expected. One of our major goals is to identify the incorrect predictions and generate a database of computationally and functionally confirmed secreted plant proteins. It has been reported that one of the most effective predictors of SPs is SignalP 3.0. We have used both SignalP and our recently developed predictor, SecreTary, to screen the predicted proteomes of several plant species to date. See below for these prediction results.