Figure 8

Similarity Score versus length. a) Linear fit between raw similarity score σ (X) (equation 8) as a function of the geometric mean $\sqrt{N_a\cdot N_b}$ of the length of the two aligned proteins (N _a and N _b are the number of residues in the two protein structures S _a and S _b ). The linear regression line (grey line) has a slope of 0.314. b) Linear fit of the normalized similarity score $\tilde{\sigma }$ (X) (equation 9) as a function of the geometric mean of the length of the two aligned proteins. The linear regression line (grey line) has a slope of -0.0004.