Update README.md
Browse files
README.md
CHANGED
@@ -26,10 +26,10 @@ SignalSeeker is a machine learning ensemble for predicting signal peptides in pr
|
|
26 |
|
27 |
## Model Performance
|
28 |
|
29 |
-
- **Best Model**:
|
30 |
-
- **Test AUC**: 0.
|
31 |
-
- **Training Data**:
|
32 |
-
- **Test Data**:
|
33 |
|
34 |
## Intended Use
|
35 |
|
@@ -99,7 +99,10 @@ The SignalSeeker ensemble consists of:
|
|
99 |
|
100 |
| Model | CV AUC | Test AUC | Test Accuracy |
|
101 |
|-------|--------|----------|---------------|
|
102 |
-
|
|
|
|
|
|
|
|
103 |
|
104 |
## Limitations
|
105 |
|
@@ -120,17 +123,17 @@ The SignalSeeker ensemble consists of:
|
|
120 |
If you use SignalSeeker in your research, please cite:
|
121 |
|
122 |
```bibtex
|
123 |
-
@misc{
|
124 |
title={SignalSeeker: Machine Learning Ensemble for Protein Signal Peptide Prediction},
|
125 |
-
author={
|
126 |
-
year={
|
127 |
-
url={https://huggingface.co/
|
128 |
}
|
129 |
```
|
130 |
|
131 |
## Contact
|
132 |
|
133 |
-
For questions or issues, please open an issue on the [GitHub repository](https://github.com/
|
134 |
|
135 |
## License
|
136 |
|
|
|
26 |
|
27 |
## Model Performance
|
28 |
|
29 |
+
- **Best Model**: Logistic regression (L2)
|
30 |
+
- **Test AUC**: 0.99433
|
31 |
+
- **Training Data**: 5000 mixed seqeunces from UniProt verified eukaryotic proteins
|
32 |
+
- **Test Data**: 1000 mixed seqeunces from UniProt verified eukaryotic proteins, isolated from training data
|
33 |
|
34 |
## Intended Use
|
35 |
|
|
|
99 |
|
100 |
| Model | CV AUC | Test AUC | Test Accuracy |
|
101 |
|-------|--------|----------|---------------|
|
102 |
+
| Logistic regression (L2) | 0.99433 | 0.98432 | 0.92284 |
|
103 |
+
| Random Forest (Regularised) | 0.98941 | 0.98869 | 0.96192 |
|
104 |
+
| Extra Trees (Regularised) | 0.99032 | 0.99072 | 0.94899 |
|
105 |
+
| SVM (Conservative) | 0.98711 | 0.98439 | 0.92284 |
|
106 |
|
107 |
## Limitations
|
108 |
|
|
|
123 |
If you use SignalSeeker in your research, please cite:
|
124 |
|
125 |
```bibtex
|
126 |
+
@misc{signalseeker2025,
|
127 |
title={SignalSeeker: Machine Learning Ensemble for Protein Signal Peptide Prediction},
|
128 |
+
author={Hugo Cooper},
|
129 |
+
year={2025},
|
130 |
+
url={https://huggingface.co/hcoops/signalseeker}
|
131 |
}
|
132 |
```
|
133 |
|
134 |
## Contact
|
135 |
|
136 |
+
For questions or issues, please open an issue on the [GitHub repository](https://github.com/hcoo25/signalseeker).
|
137 |
|
138 |
## License
|
139 |
|