(no subject)

[latentbird]

There are different techniques of shuffling. Total shuffling of all the codons will destroy local nonuniformity of amino acids composition. It is right that at the present state of research we are interested to show general meaningful of the third codon position in terms of combinatorial bias, but it seems natural that CB changes along the protein. The question is if we should use windowed shuffling in order to preserve codon distribution nonuniformity. The windowed shuffling is a shuffling within a window (say 30 codons long). The next window starts 5(10) codons further. What is a statistical prove for such a shuffling?