Figure 2From: 16S rRNA metagenome clustering and diversity estimation using locality sensitive hashingSelection of k sampled indices along with subsequence length w. Construction of randomized hash function using LSH-Div algorithm. k uniform, random indices are chosen between 1 . . . n to determine hash function h(s). For each index, w characters are chosen to the left and right to incorporate gapless subsequence matching.Back to article page