Skip to main content
Fig. 5 | BMC Bioinformatics

Fig. 5

From: OD-seq: outlier detection in multiple sequence alignments

Fig. 5

Receiver operating characteristic curves. ROC curves to verify the detectability of artificially introduced outliers in Pfam families (Table 1). It shows true and false positive values for different cutoffs with AUC of 0.978, 0.96 and 0.936 for datasets where outliers from different clans have been introduced. When introduced outliers are from the same clan AUC’s of 0.759, 0.745 and 0.732 are shown. The diagonal line represents random classification

Back to article page