DDI-PULearn: a positive-unlabeled learning method for large-scale prediction of drug-drug interactions

BMC Bioinformatics

Table 4 Top 25 novel DDIs predicted by the proposed method DDI-PULearn

Rank	Drug1	Drug1 Name	Drug2	Drug2 Name	Score
1	DB01137	Levofloxacin	DB01182	Propafenone	1
2	DB00512	Vancomycin	DB00704	Naltrexone	1
3	DB00704	Naltrexone	DB00783	Estradiol	1
4	DB00773	Etoposide	DB00783	Estradiol	1
5	DB01137	Levofloxacin	DB00704	Naltrexone	1
6	DB00635	Prednisone	DB00704	Naltrexone	1
7	DB00295	Morphine	DB01037	Selegiline	0.975
8	DB01050	Ibuprofen	DB00712	Flurbiprofen	0.975
9	DB00218	Moxifloxacin	DB00328	Indomethacin	0.975
10	DB00213	Pantoprazole	DB01189	Desflurane	0.975
11	DB00586	Diclofenac	DB01037	Selegiline	0.975
12	DB00398	Sorafenib	DB00445	Epirubicin	0.975
13	DB00724	Imiquimod	DB00331	Metformin	0.975
14	DB00203	Sildenafil	DB00457	Prazosin	0.975
15	DB00999	Hydrochlorothiazide	DB00880	Chlorothiazide	0.975
16	DB00635	Prednisone	DB00324	Fluorometholone	0.975
17	DB00704	Naltrexone	DB00327	Hydromorphone	0.975
18	DB00295	Morphine	DB00704	Naltrexone	0.975
19	DB00813	Fentanyl	DB01232	Saquinavir	0.975
20	DB00624	Testosterone	DB00959	Methylprednisolone	0.95
21	DB00959	Methylprednisolone	DB00550	Propylthiouracil	0.95
22	DB01193	Acebutolol	DB00264	Metoprolol	0.95
23	DB00295	Morphine	DB00674	Galantamine	0.95
24	DB01137	Levofloxacin	DB00323	Tolcapone	0.95
25	DB00591	Fluocinolone Acetonide	DB00641	Simvastatin	0.95

ISSN: 1471-2105