TABLE 5.

Performance statistics for species abundance estimates across the 140 mock communities (mean ± std. dev.).

Classifier	L1 distance	Relative error (%)	Abundance of FPs (%)	Abundance of FNs (%)
MCP	10.8 ± 10.47	11.2 ± 10.97	0.18 ± 0.45	0.20 ± 0.25
Unfiltered MCP	11.3 ± 10.98	11.9 ± 11.92	0.73 ± 0.96	0.08 ± 0.12
MCP w/ MGDB	10.0 ± 7.60	10.1 ± 6.69	0.27 ± 0.82	0.31 ± 0.78
Ganon	11.6 ± 9.90	10.8 ± 11.32	1.72 ± 2.32	0.03 ± 0.05
Kraken	14.4 ± 11.39	12.9 ± 13.53	3.42 ± 3.89	0.06 ± 0.35
Bracken	11.1 ± 10.33	10.5 ± 13.50	3.92 ± 4.53	0.05 ± 0.35
MetaCache	9.0 ± 8.05	10.0 ± 11.46	1.97 ± 3.00	0.02 ± 0.03
Centrifuge	49.0 ± 22.18	52.7 ± 24.92	2.97 ± 5.73	0.28 ± 0.32
DIAMOND-LCA	78.1 ± 9.54	68.6 ± 7.27	0.54 ± 0.53	3.05 ± 2.41
Kaiju	42.5 ± 13.48	34.6 ± 11.20	2.19 ± 2.59	0.20 ± 0.19
mOTUs	18.26 ± 11.86	37.2 ± 34.86	4.79 ± 5.76	4.58 ± 4.69
MetaPhlAn	43.3 ± 19.97	37.9 ± 40.37	3.58 ± 3.21	16.63 ± 10.82