Article Figures & Data
Figures
- FIG. 1.
Typical output of a best-match analysis. Visualization of the MIRU-VNTR copy numbers and spoligotype patterns of a test isolate and its best matches among the reference strain collection is presented along with information on the best-match reference strains, including the strain lineage, country of isolation, MIRU-VNTR type, SNP and RD types, and SpolDB4 shared type and lineage classification. Differences in genotype profiles between the test strain and strains in the reference database are highlighted in red for better recognition.
- FIG. 2.
Phylogenetic distribution of isolates with unassigned or T spoligotypes. A radial tree was constructed by using the 24-locus MIRU-VNTR typing data for the reference strains in the MIRU-VNTRplus database and 187 genotypes with unassigned or T spoligotypes from the Brussels population-based collection by using the neighbor-joining algorithm and categorical distance coefficient. The positions of Brussels genotypes with unassigned or T spoligotypes are indicated by yellow circles. Lineages containing exclusively T-spoligotype strains or lineages corresponding to well-characterized, classical clades of the Euro-American superlineage (such as the Haarlem lineage) comprising T-spoligotype strains are denoted T-specific and T-aspecific, respectively. EAI, East African-Indian; West Afri 2, West African 2 (M. africanum 2); West Afri 1, West African 1 (M. africanum 1); Tur, Turkish; BJ, Beijing; CAS, Central Asian; LAM, Latin American-Mediterranean.
- FIG. 3.
Strategy for use of phylogenetic identification. If no match is detected after an initial best-match analysis with a stringent distance cutoff of 0.17, tree-based analysis is preferred upon best-match analysis with a relaxed distance cutoff of 0.3 as a second step, because the former method minimizes the fine-tuned mismatching that can occur as an exception among strains belonging to the Euro-American superlineage (see text). NJ, neighbor-joining.
Tables
- TABLE 1.
Internal database evaluationa
Result No. (%) of isolates with the indicated result by the following method(s) with the indicated cutoff 24-locus MIRU-VNTR typing 24-locus MIRU-VNTR typing + spoligotyping Spoligotyping 0.17 0.3 0.17 0.3 0.17 0.3 Lineage confirmationb 147 (79.0) 177 (95.1) 170 (91.4) 183 (98.4) 175 (94.1) 178 (95.7) No match 39 (21.0) 7 (3.8) 16 (8.6) 1 (0.5) 4 (2.2) 0 (0.0) Conflictc 0 (0.0) 2 (1.1) 0 (0.0) 2 (1.1) 7 (3.8) 8 (4.3) ↵ a The self-consistency of the strain lineage identification was evaluated by testing the best matches of the genotyping data for the 186 isolates of the reference collection against the collection itself.
↵ b For lineage confirmation data, percentages correspond to sensitivity of lineage identification, defined here as the proportion of correct best matches identified among the test samples.
↵ c The specificity for lineage identification equals 100 − conflict (in percent). Specificity is defined here as the portion of correct best matches found among the total best matches identified.
- TABLE 2.
External evaluation with isolates whose lineages were preidentifieda
Cutoff and result No. (%) of isolates with the indicated result by: 24-locus MIRU-VNTR typing + spoligotyping 15-locus MIRU-VNTR typing + spoligotyping 24-locus MIRU-VNTR typing 15-locus MIRU-VNTR typing Cutoff of 0.17 Lineage confirmationb 319 (72.2) 278 (62.9) 240 (54.3) 160 (36.2) No match 123 (27.8) 163 (36.9) 199 (45.0) 282 (63.8) Conflictc 0 (0.0) 1 (0.2) 3 (0.7) 0 (0.0) Cutoff of 0.3 Lineage confirmationb 419 (94.8) 403 (91.2) 392 (88.7) 327 (74.0) No match 8 (1.8) 32 (7.2) 36 (8.1) 106 (24.0) Conflictc 15 (3.4) 7 (1.6) 14 (3.2) 9 (2.0) ↵ a Best-match-based analyses were performed with the 442 isolates with a preidentified lineage from the Brussels-Capital Region collection.
↵ b For lineage confirmation data, percentages correspond to sensitivity of lineage identification, defined here as the proportion of correct best matches identified among the test samples.
↵ c The specificity for lineage identification equals 100 − conflict (in percent). Specificity is defined here as the portion of correct best matches found among the total best matches identified.
- TABLE 3.
External evaluation with unknown or T-spoligotype isolatesa
Cutoff and result No. (%) of isolates with the indicated result by: 24-locus MIRU-VNTR typing + spoligotyping 15-locus MIRU-VNTR typing + spoligotyping 24-locus MIRU-VNTR typing 15-locus MIRU-VNTR typing Cutoff of 0.17 Best match 62 (33.2) 21 (11.2) 16 (8.6) 8 (4.3) No match 125 (66.8) 166 (88.8) 171 (91.4) 179 (95.7) Cutoff of 0.17 Best match 179 (95.7) 172 (92.0) 164 (87.7) 57 (30.5) No match 8 (4.3) 15 (8.0) 23 (12.3) 130 (69.5) ↵ a Best-match-based analyses were performed with the 187 unknown or T-spoligotype isolates from the Brussels-Capital Region collection.
