Fig. 4

OTU substructure revealed by mixed-technology OTU-picking. a PacBio CCS reads were truncated to the same length as the short reads (covering the same region of the 16S rRNA gene). Short reads and truncated CCS reads were clustered into hybrid OTUs using the oclust PW pipeline (1–3 % sequence similarity). For shared OTUs, full-length CCS reads were clustered again. b Venn diagrams showing OTUs that were jointly identified by the two technologies. The intersection denotes hybrid OTUs, i.e., containing at least one sequencing read from each technology. Only OTUs with 10 or more sequences were counted. c Full-length PacBio CCS reads were extracted, and OTU-specific clustering was performed at 3 % sequence similarity. The number of sub-OTUs indicates how many additional OTUs were inferred