Commit
·
d9cdfd9
1
Parent(s):
04b7709
Update README.md
Browse files
README.md
CHANGED
|
@@ -4,6 +4,21 @@ datasets:
|
|
| 4 |
---
|
| 5 |
# M2 - small CNN trained on embeddings
|
| 6 |
|
| 7 |
-
The model is trained on ProtBert
|
| 8 |
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
---
|
| 5 |
# M2 - small CNN trained on embeddings
|
| 6 |
|
| 7 |
+
The model is trained on [ProtBert-BFD](https://huggingface.co/Rostlab/prot_bert_bfd) embeddings of [knots_AF dataset](https://huggingface.co/datasets/EvaKlimentova/knots_AF)
|
| 8 |
|
| 9 |
+
Accuracy on the test set:
|
| 10 |
+
|
| 11 |
+
| | Dataset size | Unknotted set size | Accuracy | TPR | TNR |
|
| 12 |
+
|:----------------------------:|:------------:|:------------------:|:--------:|:------:|:------:|
|
| 13 |
+
| All | 39412 | 19718 | 0.9690 | 0.9569 | 0.9811 |
|
| 14 |
+
| SPOUT | 7371 | 550 | 0.9712 | 0.9815 | 0.8436 |
|
| 15 |
+
| TDD | 612 | 24 | 0.9673 | 0.9796 | 0.6667 |
|
| 16 |
+
| DUF | 716 | 429 | 0.9413 | 0.8955 | 0.9720 |
|
| 17 |
+
| AdoMet synthase | 1794 | 240 | 0.9727 | 0.9755 | 0.9542 |
|
| 18 |
+
| Carbonic anhydrase | 1531 | 539 | 0.8870 | 0.8619 | 0.9332 |
|
| 19 |
+
| UCH | 477 | 125 | 0.8700 | 0.8892 | 0.816 |
|
| 20 |
+
| ATCase/OTCase | 3799 | 3352 | 0.9932 | 0.9418 | 1.0 |
|
| 21 |
+
| ribosomal-mitochondrial | 147 | 41 | 0.8163 | 0.8319 | 0.7805 |
|
| 22 |
+
| membrane | 8309 | 1577 | 0.9740 | 0.9857 | 0.9239 |
|
| 23 |
+
| VIT | 14347 | 12639 | 0.9742 | 0.8214 | 0.9948 |
|
| 24 |
+
| biosynthesis of lantibiotics | 392 | 286 | 0.9388 | 0.8019 | 0.9895 |
|