7. A hanging tree as the result of hierarchical clustering of 20 amino acids.

wards, five merges happened to the amino acids N, E, Q, H, I, L,

nd T. The data set became

ൌ൫C, D, F, G, K, P, S, W, ωୖଢ଼, ω୒୉, ω୕ୌ, ω୍୐, ω୑୚, ω୅୘൯

eventh merge happened to the mean descriptor vector of the

ids R and Y as well as the amino acid K. The mean descriptor

the amino acids R and Y (ωୖଢ଼) as well as the amino acid K were

from ࣞ. In addition, the mean descriptor vector of the amino

Y and K, which was ωୖଢ଼୏, was inserted into the data set. This

ad the following structure,

ࣞൌ൫C, D, F, G, P, S, W, ωୖଢ଼୏, ω୒୉, ω୕ୌ, ω୍୐, ω୑୚, ω୅୘൯

ast merge happened to the following data set

ࣞൌ൫ω୛୊୍୐, ωୋୗ୕ୌ୅୘୒୉ୈ୔୏ୖଢ଼େ୑୚൯

y, a hanging tree was formed (Figure 2.17). The R function for

g the hierarchical cluster analysis algorithm is hclust. Its

shown below, where x is a matrix of encoded amino acids, dist

nction which generates a pairwise distance matrix between the

amino acids,

hclust(dist(x))

hclust function has the following main outputs which can be

further analysis. The plot function was called to visualise a

cal cluster model (Figure 2.17) generated by the hclust