Skip to main content
. 2011 May 4;27(12):1603–1609. doi: 10.1093/bioinformatics/btr257

Table 1.

Median scores

A C G T U
ED − 1.046 − 1.080 − 1.080 − 1.046 − 0.548
PCC 0.243 0.194 0.194 0.243 0.674
SW 0.907 0.834 0.834 0.907 1.699
KLS − 0.967 − 1.040 − 1.040 − 0.967 − 0.269
ALLR − 1.465 − 1.640 − 1.640 − 1.465 − 0.419

The median null column scores of an all-A, all-C, all-G, all-T and a uniform column. The null distribution was generated using Tomtom by scoring, using the specified column score, the alignment of the corresponding column against every column in TRANSFAC (as well as its reverse complement). The median null score of the uninformative column (U, bold) is higher than that of any of the informative columns. ED stands for Euclidean distance, PCC for Pearson's correlation coefficient (Section 7.2 in Supplementary Material), SW refers to the Sandelin–Wasserman score (Section 7.4 in Supplementary Material), KLS to the symmetric Kullback–Leibler divergence (Section 7.5 in Supplementary Material) and ALLR to the average log-likelihood ratio (Section 7.3 in Supplementary Material).