Add SST2 binding affinity results
Browse files- .gitattributes +1 -0
- README.md +19 -8
- assets/tsne_sst2_splits.png +3 -0
.gitattributes
CHANGED
|
@@ -36,3 +36,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 36 |
assets/HELM-BERT.png filter=lfs diff=lfs merge=lfs -text
|
| 37 |
assets/tsne_ppi_splits.png filter=lfs diff=lfs merge=lfs -text
|
| 38 |
assets/tsne_permeability_splits.png filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 36 |
assets/HELM-BERT.png filter=lfs diff=lfs merge=lfs -text
|
| 37 |
assets/tsne_ppi_splits.png filter=lfs diff=lfs merge=lfs -text
|
| 38 |
assets/tsne_permeability_splits.png filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
assets/tsne_sst2_splits.png filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -75,17 +75,17 @@ Pre-trained on deduplicated peptide sequences from:
|
|
| 75 |
|
| 76 |
| Split | R² | Pearson | RMSE | MAE |
|
| 77 |
|:-----:|:--:|:-------:|:----:|:---:|
|
| 78 |
-
| Random | 0.
|
| 79 |
-
| Scaffold | 0.
|
| 80 |
|
| 81 |
**Multi-Assay** (separate PAMPA and Caco-2 heads):
|
| 82 |
|
| 83 |
| Split | Assay | R² | Pearson | RMSE | MAE |
|
| 84 |
|:-----:|:-----:|:--:|:-------:|:----:|:---:|
|
| 85 |
-
| Random | PAMPA | 0.
|
| 86 |
-
| Random | Caco-2 | 0.
|
| 87 |
-
| Scaffold | PAMPA | 0.
|
| 88 |
-
| Scaffold | Caco-2 | 0.
|
| 89 |
|
| 90 |
Train/test 9:1, val 10% from train. Scaffold split by Murcko scaffolds.
|
| 91 |
|
|
@@ -95,8 +95,8 @@ Train/test 9:1, val 10% from train. Scaffold split by Murcko scaffolds.
|
|
| 95 |
|
| 96 |
| Split | ROC-AUC | PR-AUC | F1 | MCC | Balanced Acc |
|
| 97 |
|:-----:|:-------:|:------:|:--:|:---:|:------------:|
|
| 98 |
-
| Random | 0.972 | 0.
|
| 99 |
-
| aCSM | 0.
|
| 100 |
|
| 101 |
Train/test 8:2, val 10% from train, 1:4 positive:negative ratio.
|
| 102 |
- **Random**: random split
|
|
@@ -104,6 +104,17 @@ Train/test 8:2, val 10% from train, 1:4 positive:negative ratio.
|
|
| 104 |
|
| 105 |
<p align="center"><img src="assets/tsne_ppi_splits.png" width="800"></p>
|
| 106 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 107 |
## Citation
|
| 108 |
|
| 109 |
```bibtex
|
|
|
|
| 75 |
|
| 76 |
| Split | R² | Pearson | RMSE | MAE |
|
| 77 |
|:-----:|:--:|:-------:|:----:|:---:|
|
| 78 |
+
| Random | 0.769 | 0.878 | 0.388 | 0.269 |
|
| 79 |
+
| Scaffold | 0.643 | 0.812 | 0.380 | 0.284 |
|
| 80 |
|
| 81 |
**Multi-Assay** (separate PAMPA and Caco-2 heads):
|
| 82 |
|
| 83 |
| Split | Assay | R² | Pearson | RMSE | MAE |
|
| 84 |
|:-----:|:-----:|:--:|:-------:|:----:|:---:|
|
| 85 |
+
| Random | PAMPA | 0.711 | 0.844 | 0.426 | 0.298 |
|
| 86 |
+
| Random | Caco-2 | 0.772 | 0.878 | 0.402 | 0.305 |
|
| 87 |
+
| Scaffold | PAMPA | 0.584 | 0.788 | 0.393 | 0.299 |
|
| 88 |
+
| Scaffold | Caco-2 | 0.701 | 0.846 | 0.381 | 0.287 |
|
| 89 |
|
| 90 |
Train/test 9:1, val 10% from train. Scaffold split by Murcko scaffolds.
|
| 91 |
|
|
|
|
| 95 |
|
| 96 |
| Split | ROC-AUC | PR-AUC | F1 | MCC | Balanced Acc |
|
| 97 |
|:-----:|:-------:|:------:|:--:|:---:|:------------:|
|
| 98 |
+
| Random | 0.972 | 0.912 | 0.859 | 0.824 | 0.911 |
|
| 99 |
+
| aCSM | 0.868 | 0.702 | 0.613 | 0.559 | 0.735 |
|
| 100 |
|
| 101 |
Train/test 8:2, val 10% from train, 1:4 positive:negative ratio.
|
| 102 |
- **Random**: random split
|
|
|
|
| 104 |
|
| 105 |
<p align="center"><img src="assets/tsne_ppi_splits.png" width="800"></p>
|
| 106 |
|
| 107 |
+
### SST2 Binding Affinity (pChEMBL)
|
| 108 |
+
|
| 109 |
+
| Split | R² | Pearson | RMSE | MAE |
|
| 110 |
+
|:-----:|:--:|:-------:|:----:|:---:|
|
| 111 |
+
| Random | 0.312 | 0.600 | 0.742 | 0.499 |
|
| 112 |
+
| Scaffold | 0.078 | 0.532 | 1.154 | 0.821 |
|
| 113 |
+
|
| 114 |
+
Train/test 9:1, val 10% from train. Scaffold split by Murcko scaffolds.
|
| 115 |
+
|
| 116 |
+
<p align="center"><img src="assets/tsne_sst2_splits.png" width="800"></p>
|
| 117 |
+
|
| 118 |
## Citation
|
| 119 |
|
| 120 |
```bibtex
|
assets/tsne_sst2_splits.png
ADDED
|
Git LFS Details
|