Skip to main content
. 2019 Jan 8;26(2):147–156. doi: 10.1093/dnares/dsy045

Table 1.

Taxonomic classification results obtained with the proposed OTUX approach and CA

Simulated microbiomes Classification accuracies
 Target V-region
V1V2
 V1V3
 V2V3
 V3V4
 V3V5
 V3V6
 V4
 V4V5
 V4V6
 V5V6
 V6V8
Comparison of taxonomic assignmentsa at OTU level (out of 10,000 reads; averaged for 100 randomly generated simulated datasets)
GUT OTUX Correct assignments 4386.8 4355.2 5461.2 6265.4 6600.5 6957.4 5795.3 6869.8 6849.7 5862.2 6771.3
Wrong assignments 47.2 84.3 102.5 151.9 229.2 97.4 175.9 57.4 78.5 110.1 40.3
W-Unifracb from GS 0.324 0.292 0.234 0.236 0.213 0.202 0.288 0.216 0.207 0.236 0.215
CA Correct assignments 3618.6 4464.1 3808.6 3084.6 3759.3 4748.2 2318.5 2872.7 4162.0 3084.7 3429.6
Wrong assignments 555.2 244.4 1020.9 1582.2 935.7 490.4 2631.3 814.5 531.8 1785.7 1154.7
W-Unifracb from GS 0.318 0.272 0.280 0.310 0.289 0.238 0.332 0.361 0.270 0.300 0.311
SKIN OTUX Correct assignments 4034.4 4257.3 5377.9 5431.8 5710.8 5759.2 3800.5 5027.3 5800.2 5472.8 5295.9
Wrong assignments 52.7 44.9 59.8 163.7 47.5 52.6 34.0 38.3 46.8 52.9 63.8
W-Unifracb from GS 0.218 0.193 0.101 0.123 0.129 0.115 0.237 0.118 0.125 0.104 0.104
CA Correct assignments 1330.5 1789.3 1807.7 1478.2 2254.9 3238.6 320.4 1172.1 2164.7 1491.7 1816.6
Wrong assignments 1508.7 625.9 1477.0 1965.6 1476.7 727.5 1652.0 1466.1 1460.9 2346.1 1692.2
W-Unifracb from GS 0.402 0.388 0.345 0.349 0.317 0.266 0.469 0.387 0.319 0.333 0.336
SEA OTUX Correct assignments 5017.8 6085.1 6348.9 6290.1 6297.6 7037.4 4932.7 6495.9 6652.1 6253.7 6695.5
Wrong assignments 18.1 11.0 24.0 77.8 9.2 16.7 1.7 7.6 15.9 13.3 32.6
W-Unifracb from GS 0.249 0.152 0.121 0.127 0.145 0.145 0.249 0.146 0.116 0.155 0.130
CA Correct assignments 2579.8 3471.2 3180.7 3287.6 4562.6 4753.3 2183.1 3197.9 4240.4 2866.6 4047.9
Wrong assignments 1450.0 1638.7 1415.3 1174.7 1108.1 1049.9 1531.6 925.8 1095.9 1644.5 1801.0
W-Unifracb from GS 0.385 0.316 0.333 0.319 0.224 0.218 0.397 0.328 0.256 0.335 0.256
SOIL OTUX Correct assignments 5109.9 5086.0 5540.1 5431.8 5948.4 5800.0 4155.5 5257.0 5613.2 5222.1 5814.4
Wrong assignments 46.5 45.4 54.6 225.5 69.0 51.4 68.3 71.5 47.7 45.5 85.7
W-Unifracb from GS 0.145 0.114 0.078 0.123 0.127 0.153 0.225 0.133 0.120 0.162 0.125
CA Correct assignments 4089.1 4680.5 4703.2 1478.2 5176.7 5637.2 2287.0 3806.7 5357.6 3714.7 4388.2
Wrong assignments 690.0 248.5 191.5 444.4 301.4 129.3 685.9 806.1 279.4 833.5 390.7
W-Unifracb from GS 0.187 0.158 0.154 0.349 0.104 0.084 0.334 0.184 0.089 0.193 0.156

Performance of both the approaches have been evaluated with multiple simulated microbiome datasets mimicking different types of environmental samples, (viz., infant gut, skin, sea and soil), and amplicon sequencing experiments targeting different V-regions of the bacterial 16S rRNA gene. Cases wherein the number of correct assignments by OTUX are significantly higher (t-test; P < 0.001) have been highlighted in bold.

a

CA and OTUX approaches have been evaluated considering results obtained with full-length 16S rRNA genes to be ‘correct’ or the ‘GS’. Only number of average correct assignments and wrong assignments has been depicted. The remaining reads (out of 10,000) could not be classified at OTU level.

b

Weighted UNIFRAC distance of taxonomic abundance profile from actual taxonomic diversity (GS: gold-standard).