JaMBW Chapter 2.2
Randomize your sequence
This program given a sequence of symbols generates
a new version of it consisting of the same symbols but in a randomized order. Its use is mainly in conjunction of database searching
strategies, and is devoted to the evaluation of the non-randomness of a given alignment or database hit. Infact, this problem is very
often encountered when looking for distantly related sequences. The problem is then to assaying if the hit that is encountered or the
alignment that is made is bearing "real" information or if it is random or just related to the composition of that macromolecule hit.
Therefore the randomization of the same sequence is of extreme help in order to then repeat the analysis and compare the so obtained results.
would the same hit appear again on the top of the list as best alignment, would be an indication for a "randomness" of its meaning. Unfortunately
the verification for "non randomness of a distantly related hit" is rarely performed, and this could lead to fairly unstable assertions.