Table 2.

Intersection of predicted elements with the systematically identified elements reported in Table 1.

	RESCUE- ESEs [7]	Wang et al. decamers [24]	Yeo et al. 5'SS ISEs 5-mers [10]	Yeo et al. 3'SS ISEs 5-mers [10]	Zhang et al. PESEs [8]	Zhang et al. PESSs [8]	Zhang et. al. EIEs [12]	Zhang et. al. IIEs [12]	Wang et.al. ISEs/ISSs [14]	Goren et. al. ESRs [11]
5'SS ISEs			3/9.87		8/202.52		118/330.24	450/206.73	16/83.80
5'SS ISSs		105/9.46				68/54.19			46/27.40
5'SS ESEs	3/2.90				4/5.75		8/13.81	14/8.64		2/3.47
5'SS ESSs		4/0.84				0/4.61	38/22.64	19/14.17		4/5.70
3'SS ISEs				0/30.64	183/173.03		662/614.92	337/384.94	422/156.04
3'SS ISSs		156/34.31		25/35.42		83/190.50			213/164.59
3'SS ESEs	32/8.89				2/29.87		68/42.25	28/26.45		13/10.65
3'SS ESSs		0/0.21				1/1.10	19/12.98	6/8.12		5/3.27

Here is shown the ratio between the actual intersection and the expected intersection of the sets under the null hypothesis (randomly generated oligos). An intersection between the two sets of elements is calculated as the number of all the possible longest common substrings LCS of the two compared elements a and b, with the size | LCS| ≡ min(|a|, |b|), in ordered pairs (a, b) coming from the Cartesian product of the sets.