Skip to main content
Fig. 2 | BMC Bioinformatics

Fig. 2

From: Enhancing predictions of antimicrobial resistance of pathogens by expanding the potential resistance gene repertoire using a pan-genome-based feature selection approach

Fig. 2

Boxplots indicating prediction accuracies and numbers of genes of different gene sets for antimicrobial resistance (AMR) prediction problems. A The AMR prediction accuracies of different gene sets were evaluated in terms of the area under the receiver operating characteristic (AUROC) curve. B The number of genes used in predicting AMR activities for different gene sets. The gene sets included: (1) all genes; (2) known AMR genes selected using CARD/RGI; (3) genes predicted by Scoary; (4) all genes selected using the XGBoost feature selection; (5) genes selected by the incremental approach on top of the XGBoost-selected genes; and (6) the combined set of genes selected by the incremental method and known AMR genes. The y-axis is the prediction accuracies in terms of the AUROC curve

Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us