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: 94.62%	auROC: 0.995	 auPRC: 0.346
Recall at 5% | 10% | 25% | 50% FDR: 0.0% | 0.0% | 0.0% | 22.3%	Num Positives: 1135	 Num Negatives: 1022650


New best auPRC. Saving model.


Epoch 2:
Metrics across all datasets:
Balanced Accuracy: 91.73%	auROC: 0.990	 auPRC: 0.269
Recall at 5% | 10% | 25% | 50% FDR: 0.0% | 0.0% | 0.0% | 0.0%	Num Positives: 1135	 Num Negatives: 1022650



Epoch 3:
Metrics across all datasets:
Balanced Accuracy: 89.27%	auROC: 0.990	 auPRC: 0.313
Recall at 5% | 10% | 25% | 50% FDR: 0.4% | 0.4% | 0.7% | 11.7%	Num Positives: 1135	 Num Negatives: 1022650



Epoch 4:
Metrics across all datasets:
Balanced Accuracy: 86.28%	auROC: 0.986	 auPRC: 0.310
Recall at 5% | 10% | 25% | 50% FDR: 0.3% | 0.3% | 0.3% | 5.6%	Num Positives: 1135	 Num Negatives: 1022650



Epoch 5:
Metrics across all datasets:
Balanced Accuracy: 87.31%	auROC: 0.989	 auPRC: 0.325
Recall at 5% | 10% | 25% | 50% FDR: 0.0% | 0.0% | 0.0% | 18.8%	Num Positives: 1135	 Num Negatives: 1022650



Epoch 6:
Metrics across all datasets:
Balanced Accuracy: 84.11%	auROC: 0.991	 auPRC: 0.319
Recall at 5% | 10% | 25% | 50% FDR: 0.4% | 0.4% | 0.5% | 15.3%	Num Positives: 1135	 Num Negatives: 1022650


Finished training after 6 epochs.
The best model's architecture and weights (from epoch 1) were saved to /users/marinovg/2018-01-25-histone-mods/tfdragonn-SequenceAndDnaseClassifier-K562-DNAse-Duke-cuts-SP1/model.arch.json and /users/marinovg/2018-01-25-histone-mods/tfdragonn-SequenceAndDnaseClassifier-K562-DNAse-Duke-cuts-SP1/model.weights.h5