Fig. 6From: Structure-based Markov random field model for representing evolutionary constraints on functional sitesDependence of prediction on the number of aligned sequences in the multiple sequence alignment (MSA). For 47 target proteins, a the precisions, determined as the fractions of catalytic sites, at the rank cutoff of a top-L/20 and b top-L/10 (y-axis) are represented according to MSAs of different sizes (x-axis). The gray shadow represents the standard deviationBack to article page