PMC

Median percent identity of each of sequence to its nearest neighbour among the 16S rRNA cpn60 UT sequences from 1,394 bacterial genomes. Median percent identity was calculated for each 120 bp window along the length of the targets and identity values are plotted for the midpoint of each window. Due to target length variation, particularly among 16S rRNA genes, data is shown for windows for which at least 95% of the genomes could be included.

Barcode gap analysis of potential barcodes derived from the 16S rRNA and cpn60 genes. Each panel shows the distribution of inter- (red) and intra-specific (green) distances in terms of percent of the total number of comparisons made (see ). In cases where percent values exceed 20, the actual value is indicated above an arrow on the relevant bar in the chart. For both V6-alternate and cpn60 UT, only distances up to 1.00 are plotted.

A. Total error (left ordinate) for de novo assemblies of cpn60 UT sequence reads from a synthetic community of 20 cloned targets, using a minimum identity value of 92% and a range of minimum overlap lengths (50–400 nucleotides). Raw total error (blue line), as well as error remaining after post-assembly primer trimming and clustering (red line), and after chimera removal (green line). Light blue bars indicate the percent of sequence reads identified as singletons in each assembly (right ordinate). B. Number of OTU assembled at each minimum overlap length. Each coloured segment of the stacked bar indicates a different member of the panel of 20 community members. The total number of OTU assembled is indicated on the top of each stack.