FIGURE 1.

Representation of the pipeline used. Mutations are introduced into each reference genome before simulation. Simulated reads are trimmed, and human contamination removed. Reference genomes are indexed and mapped using simulated cleaned reads. Alignments are further cleaned before variant calling. All variant outputs are converted to the same format. Introduced variants covered by simulated reads are used as a set of true positives. Variants are checked in the formatted variant calls, and receiver operator characteristic (ROC) curves are calculated by repeating this process at different quality thresholds. All statistics are combined in a single file.