Figure 5.

Alu-TTS are not randomly distributed in Alu insertions and older Alu families are overrepresented. (a) Expected (red) versus observed (blue) counts of Alu-TTS are shown for individual subfamilies of different ages (older-left to younger-right). Expected counts of TTS derived from each subfamily were calculated based on the fraction of intragenic sequences. For each Alu subfamily, statistical significance levels for the differences between the expected versus observed counts (*P < 10^-4) were determined using a chi-squared distribution with df = 1. (b) Distributions of maximum utilization values (see Methods) for Alu-TTS are shown for individual subfamilies of different ages (older-left to younger-right). (c) For Alu-TTS provided by elements of different subfamilies, the position of each TTS within the subfamily consensus sequence was determined, and the density of all TTS along the length of each subfamily consensus sequence is indicated by the height of the peaks.