Skip to main content
Fig. 4 | BMC Bioinformatics

Fig. 4

From: Protein embedding based alignment

Fig. 4

Average SP scores for PEbA, vcMSA, BLOSUM62, and DEDAL as percent identity or length increases for a given reference. Pairwise identity is binned as shown in Table 4, starting from 0-9% and increasing by 10% each bin. Length is binned as shown in Table 5, starting from 0-499 and increasing by 500 each bin. References 11/12 and 911/912/913 were grouped separately due to differences in number of sequences and average sequence length in each group that resulted in varying trends across the bins. A Average SP score as percent identity increases among sequences across RV11 and RV12. B Average SP score as percent identity increases among sequences across RV911, RV912, and RV913. C Average SP score as alignment length increases among sequences across RV11 and RV12. D Average SP score as alignment length increases among sequences across BAliBASE RV911, RV912, and RV913

Back to article page