The script gets the nucleotide distribution of the input sequence list. It outputs two files. The first file shows the nucleotide distribution of the data. More precisely, a point (x, y) on the curve A shows that y sequences have x % of A.
The second plot shows the average nucleotide distribution for each position of the read. You can use it to detect a bias in the first nucleotides, for instance. A point x, y on the curve A shows that at the position x, there are y*% of **A*. A point (x, y) on the curve # tells you that y % of the sequences contain not less than x nucleotides. By definition, this latter line is a decreasing function. It usually explains why the tail of the other curves are sometimes erratic: there are few sequences.