Figure 3From: CLU: A new algorithm for EST clusteringGeneration of simulated non-matching (A) and matching (B) pairs in the process of algorithm development. A non-matching pair of sequences is generated by shuffling a randomly picked EST. A matching pair is generated by coping a stretch of 40 base pairs with one mismatch from one sequence to another. For each pair the local similarity function F(Wi) profile and categorized distribution of its values is generated.Back to article page