his represents the frequency of each score, over all the pixels, on linear (blue) and logarithmic (purple) scales; with the lowest possible score on the left and the highest on the right. The large peak on the left corresponds to the majority of pixels with low scores. If the sequences have some similarity (as is the case here), there will also be a smaller peak of higher scores. Sometimes the proportion of high-scoring pixels to the noise will be so low that the peak will be barely, if at all, perceptible on the linear scale. Hence the use of the semi-logarithmic plot.
With the scrollbars below and above the histogram, respectively, bring the lower threshold just past the first peak, and the higher threshold just past the second peak (Fig. 5).
Now, the background noise has disappeared from the dots window, and the similar regions stand out more clearly (Fig. 6):
alignment window's scrollbars have the same function