Skip to main content
. 2023 Feb 10;4(2):100691. doi: 10.1016/j.patter.2023.100691

Table 1.

Performance on the clustered split of Pfam

Method Error rate (%) Total errors
ProtCNN 27.60 5,882
ProtENN 12.20 2,590
TPHMM 18.10 3,844
BLASTp 35.90 7,639
TL-kNN 27.29 5,816
TL-MLP 19.39 4,132
TL-MLP-ensemble 18.02 3,840
TL-ProtCNN 15.98 3,405
TL-ProtCNN-ensemble∗ 8.35∗ 1,743∗

∗Model with the fewest errors.