Fig. 6

Taxonomy annotation accuracy of the DAIRYdb on reads extracted with V-Xtractor. The HVR pairs V1-V2, V2-V3, V3-V4, V4-V5, V5-V6, V6-V7, V7-V8, V8-V9 were re-annotated using three different classification algorithms, Blast+, Metaxa2 or SINTAX, respectively. This figure shows the results with SINTAX (Analyses with Metaxa2 and Blast+ are shown in Additional file 2). Taxonomy annotation was bootstrapped 1000 times with a subset of 100 randomly selected sequences from the DAIRYdb and annotated with DAIRYdb (DDB) (a), Greengenes (b), LTP (c), RDP (d) and Silva (e). Average performance of all HVR for each database (f) (accuracy = correctly annotated/total)