Background
Method
Data sources
Network type | Source database | Network details | URL address | References |
---|---|---|---|---|
DRGN | Drug bank | No. of drugs: 1497 No. of genes: 673 No. of interactions: 3509 | [14] | |
DIGN | CTD | No. of diseases: 3158 No. of genes: 47,740 No. of interactions: 26,047,815 | [15] | |
DIGN | OMIM | No. of diseases: 4552 No. of genes: 6175 No. of interactions: 6666 | [16] | |
DIGN | DisGeNET | No. of diseases: 20,371 No. of genes: 17,068 No. of interactions: 561,107 | [17] | |
PPIN | Intact | No. of proteins: 16,523 No. of interactions: 143,758 | [18] | |
GCN | COXPRESdb | No. of genes: 24,442 No. of interactions: 12,485 | [19] |
Drug–gene interaction network
Disease-gene interaction network
Protein–protein interaction network
Gene co-expression network
Reconstructing new drug-disease networks via merging heterogeneous networks
Networks | Number of drug | Number of disease | Drug-disease association |
---|---|---|---|
Net1 | 1337 | 5854 | 4,129,617 |
Net2 | 1333 | 8540 | 397,108 |
Net3 | 1191 | 10,858 | 741,819 |
Net4 | 1208 | 11,934 | 8,256,300 |
Net5 | 164 | 2240 | 82,407 |
Net6 | 239 | 2306 | 92,299 |
Net7 | 94 | 2200 | 151,267 |
Net8 | 21 | 1013 | 329 |
Net9 | 17 | 468 | 812 |
Drug-disease association prediction
Encoding drug-disease networks as feature vectors
Machine learning methods
\(Recall = \frac{TP}{TP + FN}\) | Positive correctly predicted |
\(Precision = \frac{TP}{TP + FP}\) | Positive predictive value |
\(Accuracy = \frac{{{\text{TP}} + {\text{TN}}}}{{{\text{TP}} + {\text{TN}} + {\text{FP}} + {\text{FN}}}}\) | Correctly predicted |
\(F - measure = \frac{2 \times Presion \times Sensitivity}{Presion + Sensitivity}\) | The harmonic mean of sensitivity and specificity |
-
True positive (TP): the number of drug-disease associations, which were correctly predicted.
-
True negative (TN): the number of drug-disease pairs, which were correctly predicted as non-associated.
-
False positive (FP): the number of unrelated drug-disease pairs, which were incorrectly predicted as associations.
-
False negative (FN): the number of drug-disease associations, which were incorrectly predicted as non-associations.
Benchmark dataset
Cytotoxicity assay
Results and discussion
Performance evaluation of the proposed method
Comparison with the other methods
New repurposed drugs for breast cancer
Rank | Repurposed drugs | Current usagesa | Structure |
---|---|---|---|
1 | Doxorubicin | Treatment of leukemia, lymphoma, neuroblastoma, sarcoma, Wilms tumor, and cancers of the lung, breast, stomach, ovary, thyroid, and bladder | |
2 | Paclitaxel | Treatment of AIDS-related Kaposi sarcoma, advanced ovarian cancer, and certain types of breast cancer | |
3 | Tamoxifen | Treatment of the ovary, breast cancer, desmoid tumors and endometrial cancers |