Skip to main content
Figure 4 | Journal of Cheminformatics

Figure 4

From: Benchmarking of protein descriptor sets in proteochemometric modeling (part 1): comparative study of 13 amino acid descriptor sets

Figure 4

PCA of the distances between the different descriptor sets using only the first two components per descriptor set. Shown are the first two components that explain 66% of the variance. The descriptor sets are seen to cluster based on the way they are derived. In the top left physicochemical property based descriptor sets (all derived via PCA) are shown, in the bottom left the physicochemical property based descriptor sets that are derived differently are shown. The top right shows the descriptor sets derived from topological descriptors and the bottom right shows the only molecular electrostatic potential based descriptor.

Back to article page