M1 - finetuned ProtBert-BFD
The model is trained on knots_AF dataset
The accuracy on the test set is ~ 0.9848
M1 ProtBert BFD | Dataset size | Unknotted set size | Accuracy | TPR | TNR |
---|---|---|---|---|---|
All | 39412 | 19718 | 0.9848 | 0.9871 | 0.9826 |
SPOUT | 7371 | 550 | 0.9905 | 0.9963 | 0.9182 |
TDD | 612 | 24 | 0.9918 | 0.9966 | 0.8750 |
DUF | 736 | 429 | 0.97905 | 0.9826 | 0.9767 |
AdoMet synthase | 1794 | 240 | 0.9939 | 0.9968 | 0.9750 |
Carbonic anhydrase | 1531 | 539 | 0.9556 | 0.9718 | 0.9258 |
UCH | 477 | 125 | 0.9099 | 0.9631 | 0.7600 |
ATCase/OTCase | 3799 | 3352 | 0.9992 | 0.9955 | 0.9997 |
ribosomal-mitochondrial | 147 | 41 | 0.8912 | 0.9906 | 0.63412 |
membrane | 8309 | 1577 | 0.9791 | 0.9895 | 0.9347 |
VIT | 14347 | 12639 | 0.9873 | 0.9415 | 0.9935 |
biosynthesis of lantibiotics | 392 | 286 | 0.9719 | 0.9811 | 0.9685 |
PGluconate dehydrogenase | 1 | 0 | 1.0 | 1.0 |
- Downloads last month
- 5
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.