Figure 2

Overview of the four main functionalities. A) Sequence-based feature extraction, mapping each protein in a FASTA file to a list of feature values (a row in the feature matrix). The uploaded protein labels will be used for classifier construction. B) Visual inspection of the calculated feature data, in this example showing (part of) the feature matrix in the form of a clustered heat map with in each row the feature values of one protein and the corresponding protein labels in the rightmost column. C) Classifier construction using the calculated feature matrix and the provided labels (train data). A k-fold cross-validation protocol is used to assess classification performance. D) The trained classifier can be used to predict class labels for a set of new proteins (test data).