Comparing segmentations by applying randomization techniques

BMC Bioinformatics

Table 1 Conditional entropies and their significances for segmentations on coding-noncoding data

P	H (T\|P)	p _ℓ	p _k	H (P\|T)	p _ℓ	p _k
L _cd	0.219	0	0	0.284	0	0
E _cd	0.286	0	0	0.143	0	0
L ₂	0.473	0	0	0.353	0	0
L ₁	0.566	0.0018	0.0005	0.364	0	0
L _A	0.879	0.1642	0.4219	0.466	0	0
L _C	1.212	1	0.9770	0.512	0	0.0001
L _G	0.738	0.0522	0.0732	0.849	0.1195	0.0661
L _T	0.455	0	0	0.289	0	0

Conditional entropies and their significances for segmentations on coding-noncoding data. E_cd: entropic segmentation with codon features. L_f: least-squares segmentation with features f; f ∈ {cd, 2, 1, A, C, G, T} indicate the codon feature, frequencies of 2-letter words, 1-letter words, and frequency of A, C, G, or T, respectively, p_ℓ: the fraction of segmentations from $C_{N, k, ℓ}$ with a smaller value of the conditional entropy in 10,000 randomizations; p_k: the fraction of segmentations from $C_{N, k}$ with a smaller value of the conditional entropy in 10,000 randomizations.

ISSN: 1471-2105