The pre binning assembly (PBA) cluster tree was generated by a series of clustering of PBAs at default identity cutoff for each level of taxonomy. The identity cutoffs were estimated by pairwise alignments of annotated SSU, LSU, and ITS sequences from public databases. The public taxonomy tree was mainly based on the SILVA SSU taxonomy tree and supplemented with new taxonomies from the SILVA LSU and UNITE databases. A combination of PBAs and public reference sequences (PUBs) was established based on the identities between the PBA and PUB sequences. The combined taxonomy tree contains three branch categories of sequences: 1) sequences shared by PBAs and the PUBs (red), 2) sequences unique to PBAs with no taxonomy association (blue), and 3) reference sequences from public taxonomy tree not shared by PBA (green).
