Figure 2

Three difference index structures. The numbers in right-up part of the figure gives the offset in reference sequences where the given sequence fragment have identical occurrence. In part A, blocks with dark background indicates placeholder where no actual data exists. In part B, a hash function H is performed to hash input sequences into buckets labelled with 0~4, possible conflicted sequences are chained together. Part C illustrates a binary search tree, and the number at the beginning of each block is used as the search key.