Skip to main content
Fig. 2 | Genome Biology

Fig. 2

From: Identifying tumor cells at the single-cell level using machine learning

Fig. 2

ikarus accurately classifies tumor and normal cells. A Balanced accuracy for classification of tumor and normal cells, for each of the test datasets. Red arrow highlights performance of ikarus classifier. ikarus achieved a significantly higher performance than competing methods (AUC distributions were compared using ANOVA with post hoc pairwise comparison. p values were adjusted using BH-FDR. All adjusted p values were lower than 0.01). B Area under receiver operating characteristic (AUROC) performance for each classifier. AUROC could only be calculated for methods outputting class probabilities. C Confusion matrix showing the performance of ikarus classifier on the Lambrechts lung cancer dataset. D Cell type annotation of the Lambrechts et al. dataset. E Lambrechts dataset labeled by ikarus classifier

Back to article page