Figure 4From: Protein family comparison using statistical models and predicted structural informationDistribution of profile similarity scores. (a) The uniform approach: the scores follow the EVD distribution (with λ = 0.707 and u = 4.77. (b) The normalized length-corrected scores (zscores) also follow the EVD (zscores are shifted and multiplied by 10 to emulate typical values for raw similarity scores).Back to article page