GenomeFlowInterface Settings:
shuffle: True
pos_sampling_rate: 0.05
validation_chroms: ['chr22']
holdout_chroms: ['chr1', 'chr8', 'chr21']
optimizer: adam
learning rate: 0.0003
batch size: 256
epoch size: 2500000
max num of epochs: 100
early stopping metrics: auPRC
early stopping patience: 4

Epoch 1:
Metrics across all datasets:
Balanced Accuracy: 67.70%	auROC: 0.922	 auPRC: 0.342
Recall at 5% | 10% | 25% | 50% FDR: 0.4% | 0.6% | 1.0% | 34.7%	Num Positives: 1785	 Num Negatives: 1012540


New best auPRC. Saving model.


Epoch 2:
Metrics across all datasets:
Balanced Accuracy: 92.69%	auROC: 0.985	 auPRC: 0.527
Recall at 5% | 10% | 25% | 50% FDR: 4.4% | 8.3% | 32.6% | 55.0%	Num Positives: 1785	 Num Negatives: 1012540


New best auPRC. Saving model.


Epoch 3:
Metrics across all datasets:
Balanced Accuracy: 86.45%	auROC: 0.974	 auPRC: 0.494
Recall at 5% | 10% | 25% | 50% FDR: 3.0% | 6.2% | 26.0% | 51.3%	Num Positives: 1785	 Num Negatives: 1012540



Epoch 4:
Metrics across all datasets:
Balanced Accuracy: 82.82%	auROC: 0.929	 auPRC: 0.543
Recall at 5% | 10% | 25% | 50% FDR: 5.1% | 10.8% | 38.9% | 57.4%	Num Positives: 1785	 Num Negatives: 1012540


New best auPRC. Saving model.


Epoch 5:
Metrics across all datasets:
Balanced Accuracy: 81.28%	auROC: 0.929	 auPRC: 0.508
Recall at 5% | 10% | 25% | 50% FDR: 0.1% | 7.3% | 33.4% | 51.0%	Num Positives: 1785	 Num Negatives: 1012540



Epoch 6:
Metrics across all datasets:
Balanced Accuracy: 78.29%	auROC: 0.984	 auPRC: 0.528
Recall at 5% | 10% | 25% | 50% FDR: 3.4% | 12.8% | 33.8% | 54.6%	Num Positives: 1785	 Num Negatives: 1012540



Epoch 7:
Metrics across all datasets:
Balanced Accuracy: 86.49%	auROC: 0.963	 auPRC: 0.501
Recall at 5% | 10% | 25% | 50% FDR: 0.3% | 10.9% | 32.3% | 52.8%	Num Positives: 1785	 Num Negatives: 1012540



Epoch 8:
Metrics across all datasets:
Balanced Accuracy: 72.61%	auROC: 0.964	 auPRC: 0.535
Recall at 5% | 10% | 25% | 50% FDR: 5.5% | 9.7% | 31.9% | 57.6%	Num Positives: 1785	 Num Negatives: 1012540



Epoch 9:
Metrics across all datasets:
Balanced Accuracy: 91.07%	auROC: 0.985	 auPRC: 0.462
Recall at 5% | 10% | 25% | 50% FDR: 0.0% | 0.0% | 22.0% | 48.0%	Num Positives: 1785	 Num Negatives: 1012540


Finished training after 9 epochs.
The best model's architecture and weights (from epoch 4) were saved to /users/marinovg/2018-01-25-histone-mods/tfdragonn-SequenceAndDNAseAnd11HistoneMarksClassifier_Separate3-K562-DNAse-Duke-cuts-11HistoneMarks-GATA2/model.arch.json and /users/marinovg/2018-01-25-histone-mods/tfdragonn-SequenceAndDNAseAnd11HistoneMarksClassifier_Separate3-K562-DNAse-Duke-cuts-11HistoneMarks-GATA2/model.weights.h5