Fig. 2From: PlasForest: a homology-based random forest classifier for plasmid detection in genomic datasetsChosen features and their importances in the classification process. A Schematic representation of the features extracted from contigs, including homology-based features (number of hits, maximum overlap, average overlap, median overlap, variance of overlaps, contig size) and sequence-based feature (G + C content). B Impurity-based feature importance computed with scikit-learn library for the seven features kept in the classifierBack to article page