Fig. 4From: TaxMapper: an analysis tool, reference database and workflow for metatranscriptome analysis of eukaryotic microorganismsClassification scheme. One million reads from different taxonomic groups with 50% false positive and 50% true positive best hit assignments were used. This dataset was split in 20% holdout data and 80% training and test data, of which again 80% were used to train and 20% to test the classifier applying 100-fold Monte Carlo cross-validation. In addition, in each fold the holdout data and randomly simulated (nonsense) reads were used to evaluate the classifierBack to article page