Figure 3
From: Optimal neighborhood indexing for protein similarity search
Neighborhood indexing. Fragment of a neighborhood index. For each seed key, the list of its occurrence positions is stored. For each occurrence, its right and left neighborhoods are additionally stored.