Fig. 4From: geneRFinder: gene finding in distinct metagenomic data complexitiesModels trained on a training set with 712,886 sequences. Random Forest trees with 100 models are less complex and robust than large ensembles trees of 700 modelsBack to article page