Table 2. Pyrosequencing metrics of the cleaned data and its distribution at the Kingdom level.

Horse	After cleaning	After filters	Bacteria	Eukaryota	Unassigned	Unclassified seq.	Archaea
Healthy 1	7232	6948	90.27	6.47	3.14	0.12	–
Healthy 2	5622	5351	90.92	7.12	1.89	0.07	–
Healthy 3	17725	19328	88.83	7.32	3.62	0.22	0.01
Healthy 4	13202	11270	89.59	9.74	0.24	0.43	–
Healthy 5	4962	4402	78.10	13.22	–	8.61	0.07
Healthy 6	12872	12201	80.60	13.91	0.98	4.52	–
Mean (±SD)	10,269(5,098)	9,917(5,580)	86.85	9.45	1.96	1.74	0.01
Colitis 1	6259	5146	92.93	3.67	2.60	0.80	–
Colitis 2	5257	2996	98.10	0.53	1.37	–	–
Colitis 3	5219	4245	84.31	15.19	0.19	0.31	–
Colitis 4	8134	7031	72.29	27.04	0.67	–	–
Colitis 5	10669	9878	77.70	3.76	17.61	0.93	–
Colitis 6	30285	33206	84.73	11.00	4.16	0.11	–
Colitis 7	6952	7085	99.15	0.20	0.07	0.58	–
Colitis 8	12508	11910	87.36	8.25	3.38	1.02	–
Colitis 9	7613	7493	96.68	2.87	0.32	0.13	–
Colitis 10	8275	7929	86.91	5.99	6.96	0.14	–
Mean (±SD)	10,117(7,445)	9,692(8,659)	86.42	8.73	4.47	0.38	0.00
Total	162,786	156,419

Total number of reads after data cleaning (pyrosequencing noise and chimera removal), after filtering (e-value of 30, minimum identity of 97% and minimum alignment of 75bp on MG-RAST), and percentage of reads classified by MG-RAST using the SSU databank as Bacteria, Eukaryota, Archaea, unclassified bacteria and sequences unassigned to any Kingdom. Means and standard deviations (±SD) among healthy horses and horses with colitis are also presented.