Table 3.

Events Detected in the Data Sets

	E. coli
	CRISPR1.1	CRISPR1.2	CRISPR2.1	CRISPR2.2	P. aeruginosa	S. agalactiae	S. thermophilus
Arrays	1,302	1,302	142	142	289	243	84
Unique arrays	356	345	41	44	198	210	70
Median length	10	9	9	11	13	10.5	23
Unique spacers	746	766	146	140	1,315	750	896
Long spacers (≥100 nt)	6	1	0	0	5	0	2
Order inversions
Replicated spacers	35 (4.7%)	35 (4.6%)	3 (2.1%)	15 (11%)	19 (1.4%)	30 (4%)	35 (3.9%)
Others	7 (0.94%)	2 (0.26%)	0	0	0	15 (2%)	0
Deletions
Deletion events	257	316	25	28	134	159	34
Unique deleted spacers	278 (37%)	314 (41%)	31 (21%)	43 (31%)	280 (21%)	158 (21%)	120 (13%)
ODEs	40	46	0	4	40	30	15
Multiple-spacer ODEs	26	20	0	3	17	19	9
Min. number of spacers in events	45 (6.1%)	58.5 (7.6%)	0	7 (0.5%)	45 (3.4%)	35 (4.7%)	27 (3.0%)
Max. number of spacers in events	170.2 (23%)	220 (29%)	0	19 (14%)	309 (23%)	128 (17%)	120 (13%)
Max. number of spacers in multiple-spacer events	129.5 (17%)	109 (14%)	0	17 (12%)	112 (8.5%)	80 (11%)	71 (7.9%)
Ratio ODEs to deletion events	0.1556	0.1456	0	0.1429	0.2985	0.1887	0.4412
Ratio multiple-spacer ODEs to deletion events	0.1012	0.0633	0	0.1071	0.1269	0.1195	0.2647
Estimation of deletion effects
Artificial to simulated deletion events	0.2574	0.2192	0.1997	0.1165	0.2909	0.121	0.2989
Proximal deletions to deletion events	0.2084	0.2389	0.1169	0.3991	0.1947	0.3997	0.1234
Multiple-spacer proximal deletions to deletion events	0.156	0.1497	0.0944	0.2275	0.1258	0.2083	0.1105

Note.—Estimation of deletion effects: Ratios are given using the median for 1–30 deletions, except for E. coli CRISPR2 and S. thermophilus. There only up to 15 deletions are used, because the ratio of inferred to simulated deletions is decreasing with high numbers of deletions for these data sets (data not shown).