Fig. 3

Choice of assembler causally influences pseudogene counts. Cumulative distributions for frameshifts per Mbp (top left) and internal stops per Mbp (top right) for reassembled N. gonorrhoeae genomes with available Illumina reads (n = 242). Pseudogenes were matched across the four reassemblies of each genome at ≥ 90% similarity with Clusterize. Venn Diagrams show the overlap in frameshifts (bottom left) and internal stops (bottom right) for the reassembled genomes. Overall, only about 70% of pseudogenes were shared by all four assemblers