Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: ProfNet, a method to derive profile-profile alignment scoring functions that improves the alignments of distantly related proteins

Figure 1

The distribution of scores from family, superfamily, fold related and randomly chosen profile vectors for prob_score and the five different ProfNet versions. The ProfNet versions were trained on profile vector pairs from unrelated proteins and proteins related at family (ProfNet_fam), superfamily (ProfNet_su), fold (ProfNet_fold), and a combination of family, superfamily and fold (ProfNet_all). The S-score training (ProfNet_S) was done using superfamily related vector pairs as positive examples, and classified by the S-score instead of the binary classifiers used in the other cases. All graphs show a Gaussian distribution, except for the family related scores in ProfNet_fam, which instead seems to follow an extreme-value like distribution. In each plot, the fraction of residues within a certain score range is plotted against the score. The exact values of the Y-axis have been left out for clarity.

Back to article page