BBBP prediction

(Blood-Brain-Barrier-Permeability prediction)

Prepared datasets

Train set	Test sets
train_set.csv (LightBBB train set)	test_set.csv (LightBBB test set) | test_set_6.csv (B3DB dataset with logBB)
train_set_2.csv (MoleculeNet)	test_set.csv (LightBBB test set) | test_set_combined.csv (LightBBB train_test set) | test_set_2.csv (LightBBB test set more) | test_set_3.csv (B3DB dataset) | test_set_4.csv (B3DB dataset with logBB) | test_set_5.csv (B3DB dataset with |logBB| >=1)

Available trained model:

Balanced Random Forest Classifier

Balanced RF Classifier	model/RF_model.pickle.dat
Number of features	743 (model/RF_features.pickle)
Number of estimators	400
Train set	dataset/feature_set_all/train_set_2.csv

Performance

1. Test_set.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	16	16	1	0	0.97	0.94	1.00	0.97	0.97	0.94
LightBBB	16	13	1	3	0.88	0.93	0.81	0.87	0.88	0.85

2. Test_set_combined.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	1452	683	189	239	0.83	0.88	0.86	0.87	0.82	0.85

3. Test_set_2.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	17	23	2	2	0.91	0.89	0.89	0.89	0.91	0.85

4. Test_set_3.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	3001	2044	482	556	0.83	0.86	0.84	0.85	0.83	0.82

5. Test_set_4.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	518	65	34	154	0.76	0.94	0.77	0.85	0.71	0.92

6. Test_set_5.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	48	65	34	10	0.72	0.59	0.83	0.69	0.74	0.55

7. Test_set_6.csv

Model	tp	tn	fp	fn	accuracy	precision	recall	f1	roc_auc	avg_precision
Balanced RF	179	36	14	84	0.69	0.93	0.68	0.79	0.70	0.90
LightBBB	156	28	22	107	0.59	0.88	0.59	0.71	0.58	0.86