Figure 5

Sensitivity and specificity of the taxonomic classification programs using custom databases. The sensitivity and the specificity were calculated based on the results of taxonomic classification trials using three custom databases in which genomes with taxonomic labels matching those in the query dataset were respectively removed: Species-excluded (A), Genera-excluded (B) and Families-excluded (C). The query dataset, comprised of the 200 simulated reads, was analysed with three taxonomic classification approaches with each database: K-mer frequencies, tBLASTx, and PhymmBL. These methods were assessed at the taxonomic levels of genus (blue), family (red) and order (green) whenever available for each iteration.