Figure 5From: Can single knockouts accurately single out gene functions?Clustering of tasks by participating genes. Hierarchical clustering of the 41 subtasks according to the binary vectors of participating genes from Figure 4. Clustering with a standard agglomerative, shortest single link algorithm, using a Jaccard distance metric with Matlab software. Tasks are manually classified according to product biochemical identity, to illuminate the clustering qualities: A mino acids, P urines, pY rimidines, pH ospholipids, S terols and C arbohydrates.Back to article page