Fig. 3.

The distribution of frequencies for correct variants after thresholding at 1%. Each point represents the frequency of a codon in the plasmid sequences. The overlapping codons with frequency near 1.0 represent identical sites across all plasmid sequences. At sites where the plasmid sequences differ (points not near 1.0), we expect the frequencies to follow the mixture proportions of the plasmid sequences in the dataset. Statistical tests of matched comparisons among the methods show no significant differences, except for bowtie2-pear on the 5VM dataset