Fig. 2.

Workflow for differentiating alleles from artifacts: Step II. This workflow is modified from [19] and a more detailed description of the workflow is provided in the Additional file 1. In Step II, identical reads within each amplicon were clustered into a single variant and variants were sorted according to relative frequency within the amplicon. Variants of > 1 % frequency and chimeric variants were discarded. Then variants were classified as differing from the most frequent sequence by 1–2 bp or > 2 bp. Intra-amplicon classifications are shaded in gray, whereas final variant identifications are shaded in blue. Variant numbers are given in bold