The dimer symbol frequencies counted in all proteins. Similar to symbol-distribution but counts dimers.

Example: Assume that you provided 2 sequence sets of the alphabet {+,0,-} and the first sequence in the set A would be "+++0--". Then in the first sequence these patterns are found exactly: "++", "+0", "0-" and "--". At all in our first sequence 5 patterns with 2 symbols are possible if you move over the sequence with a sliding window of 2 residues. The frequency shown is the relative fraction of the occurences of this pattern to the number of possible patterns in the respective protein.

DATASET PATERN FREQ
setA ++ 2/5 = 0.4
setA +0 1/5 = 0.2
setA 0- 0.2
setA --
0.2

Description of set A: '$SETA'
Description of set B: '$SETB'