Table 2.

Mean percentages of correct predictions at the rank of genus and at the higher ranks of 13 taxon groups

	Genus	CDS		Non-CDS		Genus	CDS		Non-CDS
		Correct genus (%)	Correct taxon group (%)	Correct genus (%)	Correct taxon group (%)		Correct genus (%)	Correct taxon group (%)	Correct genus (%)	Correct taxon group (%)
Bacteria	Acidithiobacillus	57	78	51	72	Microbacterium	87	96	60	89
	Acidobacterium	62	62	40	40	Micrococcus	93	97	48	89
	Agrobacterium	65	84	50	65	Myxococcus	88	89	27	39
	Anabaena	41	57	56	78	Nitrobacter	66	90	42	71
	Azorhizobium	87	97	49	80	Nitrosococcus	51	66	42	69
	Azotobacter	75	87	55	71	Nitrosomonas	45	45	33	33
	Bacillus	53	53	64	64	Nitrosospira	60	60	70	72
	Bdellovibrio	61	64	63	66	Nocardia	79	89	25	44
	Beijerinckia	65	83	56	66	Nostoc	58	60	62	68
	Bradyrhizobium	84	88	41	61	Oscillatoria	58	58	66	66
	Caulobacter	79	91	43	59	Pseudanabaena	76	77	48	53
	Clostridium	81	85	90	92	Pseudomonas	77	95	52	64
	Cyanobacterium	72	73	61	61	Pseudonocardia	88	96	29	74
	Desulfotomaculum	49	54	43	69	Rhizobium	70	81	48	60
	Desulfovibrio	43	52	52	55	Rhodobacter	85	94	32	66
	Erwinia	71	87	47	81	Rickettsia	68	68	67	67
	Frankia	72	90	18	50	Shewanella	75	83	83	86
	Geobacter	61	67	47	54	Sinorhizobium	67	83	51	69
	Klebsiella	79	95	59	77	Sphingomonas	76	92	31	71
	Kocuria	88	97	52	89	Streptomyces	89	96	55	56
	Leuconostoc	62	72	41	73	Variovorax	85	85	41	41
	Mesorhizobium	70	90	40	58	Xanthomonas	91	94	48	60
	Methylococcus	76	87	60	82
	Average accuracy	3185/4500	3587/4500	2238/4500	2970/4500	Mean percentage of correct predictions	71%	80%	50%	66%
Fungi	AMF	49	49	72	72	Oidiodendron	68	68	71	71
	Aspergillus	72	72	77	77	Phanerochaete	52	67	66	91
	Cenococcum	54	66	88	90	Scleroderma	76	87	68	89
	Cryptococcus	72	87	78	89	Sebacina	58	89	66	90
	Mycosphaerella	88	93	69	81
	Average accuracy	589/900	678/900	655/900	750/900	Mean percentage of correct predictions	65%	75%	73%	83%

Note: Average accuracy was calculated by dividing sum of correct predictions by total number of predictions made. For example, 3185 out of 4500 sequences were correctly classified to its taxonomic group at the rank of genus in testing bacterial CDS. Mean percentage of correct predictions was calculated by multiplying average accuracy by 100. For example, for bacterial CDS, (3185/4500) × 100 = 71. After genus in an answer was converted to a corresponding taxonomic group in the 13 taxon groups, the mean percentage of correct predictions (Correct taxon group %) was calculated. 13 taxon groups are indicated in the section, Scoring methods.