The functions implemented in RMs are replaced by the contents of the bit files.
The cluster has an average information content of 2.68 (bits), which denotes higher conservation in the sequence of these fragments.
Ignoring effects of boundaries and mutual intersection of individual block positions, the number of possible six-block patterns will be about 45 = 210, corresponding to an information content of 10 bits per pattern.
For example, F.b.50.1.3.11.7870 is a homogeneous cluster for acid proteases with information content of 3 bits and with an average distances between the two ends of 11 Å, having a ranking of 7,870.
A similar calculation for the complex pattern, where 16 positions of the second block are possible, yields 165 = 220 possible patterns corresponding to an information content of 20 bits per pattern.
Ernout's text contains N = 49,658 words belonging to L = 4,492 lemmata [33], and is found from equation (6) to have an entropy information content of I≈474 ·103 bits (∼58KB).
Our hERα PWM matrix has an information content of 16 bits ([see Additional file 8], panel B, logo and frequency matrix).
The TRANSFAC [ 35] hERα PWM matrix, 19 bases wide, is poorly polarized for the first half of the dimer ([see Additional file 8], panel A), thus corresponding to an information content of 10.3 bits.
It follows that the generally accepted maximal information content of 2 bit per base pair only holds true if the two sister strands can be distinguished (which requires extra information).
We define the 'best case' background column as having an information content of 1 bit (equivalent to a PWM column such as (0.5,0.5,0.0,0.0) T, where any two nucleotides are split equally).
