Table 4.
Uncovering the False Negatives examples in Twosides database
| Drug1 | Drug2 | Side effect (SE) | SE freq (%) | Probability | ||
|---|---|---|---|---|---|---|
| Random | One-unseen | Both-unseen | ||||
| Fenofibrate | Valsartan | Hepatitis c | 5.1 | 0.008 | – | – |
| Verapamil | Fluvoxamine | Hepatitis c | 5.1 | 0.019 | 0.034 | – |
| Bupropion | Fluoxetine | Hepatitis a | 1.0 | 0.052 | 0.008 | – |
| Bupropion | Fluoxetine | HIV disease | 0.9 | 0.084 | 0.009 | – |
| Paroxetine | Risperidone | HIV disease | 0.9 | 0.035 | 0.006 | – |
| Mupirocin | Sertraline | Drug withdrawal | 12.5 | 0.015 | 0.107 | 0.056 |
| Amlodipine | Cerivastatin | Flu | 10.4 | 0.028 | 0.050 | – |
| Metoclopramide | Minoxidil | Road traffic accident | 14.6 | 0.048 | 0.005 | – |
| n-Acetylcysteine | Cefuroxime | Herpes simplex | 6.8 | 0.029 | 0.028 | – |
| Clonazepam | Salmeterol | Adverse drug effect | 7.5 | 0.048 | – | 0.048 |
The table shows the drug–drug potential side effect and their probability of existence predicted by the 1D-CNN model