Table 3. Comparison of sensitivity and specificity.

	Sensitivity		Specificity
Method	No.	%	No.	%
Gibbs sampler (ours)	41/53	77.4	41/49	83.7
compareprospector	(29 + 1)/53	56.6	(29 + 1)/48	62.5
phylocon	(24 + 1)/53	47.2	(24 + 1)/35	71.4

The + 1s in the second and third rows mean that one motif predicted by compareprospector and one motif predicted by phylocon look similar to the corresponding experimentally verified motif in TRANSFAC, respectively, although they do not satisfy our criteria of correct prediction. The number 41/53 in the first row means Gibbs sampler (ours) correctly made 41 predictions for the total 53 datasets; the numbers (29 + 1)/53 in the second and third rows have similar meaning. The number 41/49 in the first row means Gibbs sampler (ours) correctly made 41 predictions in a total of 49 predictions; the number (29 + 1)/48 in the second row and the number (24 + 1)/35 in the third row have similar meaning.