Table 1.
Strains used in this work.
Table 2.
Strains, DTUs, NCBI SRA accession codes of the analyzed whole genome files.
Fig 1.
Strains of the same DTU shared mHVR clusters at different identity thresholds.
Similarity matrices show the number of mHVR clusters shared between strains of the same lineage and between strains of different lineages. Strains were arranged according to their lineage. The color scale indicates the similarity between pairs of strains. A, 85% identity threshold; B, 95% identity threshold. White: 0 shared mHVR clusters, yellow-orange: less than 20 shared mHVR clusters, red: more than 20 shared mHVR clusters.
Fig 2.
The usefulness of sets of mHVR reference sequences for typing each strain.
The proportion of reads clustered with the reference sequences of each DTU is shown as horizontal bars for each strain. The color bars represent the proportion of reads that clustered with the reference sequences from each DTU. At the center, the DTU to which each strain belongs is indicated. Blue bars: TcI, orange bars: TcII, gray bars: TcIII, yellow bars: TcIV, violet bars: TcV, and green bars: TcVI. Each analyzed strain was typified using a reference set that excluded the sequences of the analyzed strain. Two different groups of reference sets were tested based on the mHVR clusters constructed with 85% (left) and 95% (right) similarity thresholds.
Fig 3.
The usefulness of the 95% mHVR reference sequences set for typing data from whole-genome projects.
The whole-genome reads for different strains were mapped to mHVR reference sequences of each DTU. A- The color bars for each strain represent the percentage of mHVR reference sequences for each DTU that were successfully mapped with a coverage of 270 bases at 10X depth. B- The color bars for each strain represents the percentage of the total number of bases for the whole-genome reads mapped to the mHVR reference sequences of each lineage with a coverage of 270 bases at 10X depth. At the center, the DTU to which each strain belongs is indicated. Blue bars: TcI, orange bars: TcII, gray bars: TcIII, yellow bars: TcIV, violet bars: TcV, and green bars: TcVI.
Fig 4.
The efficiency of reference sets for typing simulated PCRs.
A, D: Average sensitivity for the reference set constructed at the 95% similarity threshold for different simulated PCR conditions. B, E: Average specificity for typing DTUs based on the most abundant DTU-tag identified, while discarding minority DTU-tags in a sample for the 95% and 85% reference sets. C, F: Average false-positive rate for DTU detection considering all DTU-tags in a sample for the 95% and 85% reference sets. PCRs were simulated with 100 starting DNA molecules randomly selected from each strain dataset and 99% efficiency (black bars), 10 starting DNA molecules randomly selected from each strain and 80% efficiency (gray bars), and one randomly selected initial DNA molecule from each strain and 70% efficiency (white bars).
Table 3.
Cutoffs for the minimum DTU-tag frequency indicate the presence of a secondary DTU in the sample according to the main DTU with the associated error probability for PCR simulations based on the 95% reference set.
Table 4.
Cutoffs for the minimum DTU-tag frequency indicate the presence of a secondary DTU in the sample according to the main DTU with the associated error probability for PCR simulations based on the 85% reference set.
Fig 5.
Sensitivity for detecting secondary DTUs in the simulated samples.
A and B: Sensitivities using 95% reference set. C and D, sensitivities using an 85% reference set. Mock samples were simulated with different proportions of the main and secondary DTU. A and C: 90% of the reads of the main DTU and 10% of the secondary DTU. B and D: 95% of the reads of the main DTU and 5% of the secondary DTU. The strains used for each DTU were PalDa20cl3 (TcI), Esmeraldo (TcII), X109/2 (TcIII), MNcl2 (TcV) and LL015P68R0cl4 (TcVI).
Fig 6.
Comparison between Southern blot and deep amplicon sequencing.
A, DTUs identified in blood samples of 28 patients by Southern blot by using mHVR probes (left) and deep amplicon sequencing by using the 95% reference set. B, comparison of the prevalence for different DTUs determined by using Southern blot (blue bars) and deep amplicon sequencing by using the 95% reference set (green bars). ND = nondetermined DTU by the method.