Performance comparison with SOTA methods. The best results are in bold font.
| Method . | Year . | Dataset . | #Drugs . | Rep. . | Model . | Aiming . | Coverage . | Accuracy . | Absolute . | Absolute . | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| . | . | . | . | . | . | . | . | . | |$\uparrow $| . | |$\uparrow $| . | |$\uparrow $| . | True |$\uparrow $| . | False |$\downarrow $| . |
| Chen et al. [4] | 2012 | Chen-2012 | 3883 | I | S | Similarity Search | 50.76% | 75.79% | 49.38% | 13.83% | 8.83% | ||
| iATC-mISF[5] | 2017 | Chen-2012 | 3883 | I | S | F | ML-GKR | 67.83% | 67.10% | 66.41% | 60.98% | 5.85% | |
| iATC-mHyb[6] | 2017 | Chen-2012 | 3883 | I | S | F | O | ML-GKR | 71.91% | 71.46% | 71.32% | 66.75% | 2.43% |
| EnsLIFT[7] | 2017 | Chen-2012 | 3883 | I | S | F | LIFT | 78.18% | 75.77% | 71.21% | 63.30% | 2.85% | |
| EnsANet_LR[9] | 2018 | Chen-2012 | 3883 | I | S | F | CNN,LIFT,RR | 75.40% | 82.49% | 75.12% | 66.68% | 2.62% | |
| EnsANet_LR|$\otimes $|DO[9] | 2018 | Chen-2012 | 3883 | I | S | F | O | CNN,LIFT,RR | 79.57% | 83.35% | 77.78% | 70.90% | 2.40% |
| ATC-NLSP[10] | 2019 | Chen-2012 | 3883 | I | S | F | NLSP | 81.35% | 79.50% | 78.28% | 74.97% | 3.43% | |
| iATC-NRAKEL[11] | 2020 | Chen-2012 | 3883 | I | S | RAKEL,SVM | 78.88% | 79.36% | 77.86% | 75.93% | 3.63% | ||
| iATC-FRAKEL[12] | 2020 | Chen-2012 | 3883 | F | RAKEL,SVM | 78.51% | 78.40% | 77.21% | 75.11% | 3.70% | |||
| FUS3[14] | 2020 | Chen-2012 | 3883 | I | S | F | CNN,LSTM,LIFT,RR | 87.55% | 69.73% | 73.46% | 68.71% | 2.38% | |
| FUS3|$\otimes $|DO[14] | 2020 | Chen-2012 | 3883 | I | S | F | O | CNN,LSTM,LIFT,RR | 79.79% | 84.22% | 79.64% | 73.04% | 2.09% |
| iATC_Deep-mISF[13] | 2020 | Chen-2012 | 3883 | I | S | F | O | DNN | 74.70% | 73.91% | 71.57% | 67.01% | 0.00% |
| CGATCPred[15] | 2021 | Chen-2012 | 3883 | I | S | E | A | CNN,GCN | 81.94% | 82.88% | 80.81% | 76.58% | 2.75% |
| EnsATC[18] | 2022 | Chen-2012 | 3883 | I | S | F | hMuLab,LSTM | 91.39% | 84.32% | 83.38% | 80.09% | 1.31% | |
| ATC-CNN | 2022 | Chen-2012 | 3883 | S | CNN | 93.01% | 90.72% | 90.53% | 87.77% | 1.53% | |||
| ATC-CNN | 2022 | ATC-SMILES | 4545 | S | CNN | 95.83% | 94.14% | 93.99% | 91.77% | 0.94% | |||
| Method . | Year . | Dataset . | #Drugs . | Rep. . | Model . | Aiming . | Coverage . | Accuracy . | Absolute . | Absolute . | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| . | . | . | . | . | . | . | . | . | |$\uparrow $| . | |$\uparrow $| . | |$\uparrow $| . | True |$\uparrow $| . | False |$\downarrow $| . |
| Chen et al. [4] | 2012 | Chen-2012 | 3883 | I | S | Similarity Search | 50.76% | 75.79% | 49.38% | 13.83% | 8.83% | ||
| iATC-mISF[5] | 2017 | Chen-2012 | 3883 | I | S | F | ML-GKR | 67.83% | 67.10% | 66.41% | 60.98% | 5.85% | |
| iATC-mHyb[6] | 2017 | Chen-2012 | 3883 | I | S | F | O | ML-GKR | 71.91% | 71.46% | 71.32% | 66.75% | 2.43% |
| EnsLIFT[7] | 2017 | Chen-2012 | 3883 | I | S | F | LIFT | 78.18% | 75.77% | 71.21% | 63.30% | 2.85% | |
| EnsANet_LR[9] | 2018 | Chen-2012 | 3883 | I | S | F | CNN,LIFT,RR | 75.40% | 82.49% | 75.12% | 66.68% | 2.62% | |
| EnsANet_LR|$\otimes $|DO[9] | 2018 | Chen-2012 | 3883 | I | S | F | O | CNN,LIFT,RR | 79.57% | 83.35% | 77.78% | 70.90% | 2.40% |
| ATC-NLSP[10] | 2019 | Chen-2012 | 3883 | I | S | F | NLSP | 81.35% | 79.50% | 78.28% | 74.97% | 3.43% | |
| iATC-NRAKEL[11] | 2020 | Chen-2012 | 3883 | I | S | RAKEL,SVM | 78.88% | 79.36% | 77.86% | 75.93% | 3.63% | ||
| iATC-FRAKEL[12] | 2020 | Chen-2012 | 3883 | F | RAKEL,SVM | 78.51% | 78.40% | 77.21% | 75.11% | 3.70% | |||
| FUS3[14] | 2020 | Chen-2012 | 3883 | I | S | F | CNN,LSTM,LIFT,RR | 87.55% | 69.73% | 73.46% | 68.71% | 2.38% | |
| FUS3|$\otimes $|DO[14] | 2020 | Chen-2012 | 3883 | I | S | F | O | CNN,LSTM,LIFT,RR | 79.79% | 84.22% | 79.64% | 73.04% | 2.09% |
| iATC_Deep-mISF[13] | 2020 | Chen-2012 | 3883 | I | S | F | O | DNN | 74.70% | 73.91% | 71.57% | 67.01% | 0.00% |
| CGATCPred[15] | 2021 | Chen-2012 | 3883 | I | S | E | A | CNN,GCN | 81.94% | 82.88% | 80.81% | 76.58% | 2.75% |
| EnsATC[18] | 2022 | Chen-2012 | 3883 | I | S | F | hMuLab,LSTM | 91.39% | 84.32% | 83.38% | 80.09% | 1.31% | |
| ATC-CNN | 2022 | Chen-2012 | 3883 | S | CNN | 93.01% | 90.72% | 90.53% | 87.77% | 1.53% | |||
| ATC-CNN | 2022 | ATC-SMILES | 4545 | S | CNN | 95.83% | 94.14% | 93.99% | 91.77% | 0.94% | |||
Representation (Rep.) abbreviations: I - Chemical interactions, S - Chemical structural features, F - Molecular fingerprint features, O - Drug ontology information, E - Pre-trained word embedding, and A - ATC codes association information.
Performance comparison with SOTA methods. The best results are in bold font.
| Method . | Year . | Dataset . | #Drugs . | Rep. . | Model . | Aiming . | Coverage . | Accuracy . | Absolute . | Absolute . | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| . | . | . | . | . | . | . | . | . | |$\uparrow $| . | |$\uparrow $| . | |$\uparrow $| . | True |$\uparrow $| . | False |$\downarrow $| . |
| Chen et al. [4] | 2012 | Chen-2012 | 3883 | I | S | Similarity Search | 50.76% | 75.79% | 49.38% | 13.83% | 8.83% | ||
| iATC-mISF[5] | 2017 | Chen-2012 | 3883 | I | S | F | ML-GKR | 67.83% | 67.10% | 66.41% | 60.98% | 5.85% | |
| iATC-mHyb[6] | 2017 | Chen-2012 | 3883 | I | S | F | O | ML-GKR | 71.91% | 71.46% | 71.32% | 66.75% | 2.43% |
| EnsLIFT[7] | 2017 | Chen-2012 | 3883 | I | S | F | LIFT | 78.18% | 75.77% | 71.21% | 63.30% | 2.85% | |
| EnsANet_LR[9] | 2018 | Chen-2012 | 3883 | I | S | F | CNN,LIFT,RR | 75.40% | 82.49% | 75.12% | 66.68% | 2.62% | |
| EnsANet_LR|$\otimes $|DO[9] | 2018 | Chen-2012 | 3883 | I | S | F | O | CNN,LIFT,RR | 79.57% | 83.35% | 77.78% | 70.90% | 2.40% |
| ATC-NLSP[10] | 2019 | Chen-2012 | 3883 | I | S | F | NLSP | 81.35% | 79.50% | 78.28% | 74.97% | 3.43% | |
| iATC-NRAKEL[11] | 2020 | Chen-2012 | 3883 | I | S | RAKEL,SVM | 78.88% | 79.36% | 77.86% | 75.93% | 3.63% | ||
| iATC-FRAKEL[12] | 2020 | Chen-2012 | 3883 | F | RAKEL,SVM | 78.51% | 78.40% | 77.21% | 75.11% | 3.70% | |||
| FUS3[14] | 2020 | Chen-2012 | 3883 | I | S | F | CNN,LSTM,LIFT,RR | 87.55% | 69.73% | 73.46% | 68.71% | 2.38% | |
| FUS3|$\otimes $|DO[14] | 2020 | Chen-2012 | 3883 | I | S | F | O | CNN,LSTM,LIFT,RR | 79.79% | 84.22% | 79.64% | 73.04% | 2.09% |
| iATC_Deep-mISF[13] | 2020 | Chen-2012 | 3883 | I | S | F | O | DNN | 74.70% | 73.91% | 71.57% | 67.01% | 0.00% |
| CGATCPred[15] | 2021 | Chen-2012 | 3883 | I | S | E | A | CNN,GCN | 81.94% | 82.88% | 80.81% | 76.58% | 2.75% |
| EnsATC[18] | 2022 | Chen-2012 | 3883 | I | S | F | hMuLab,LSTM | 91.39% | 84.32% | 83.38% | 80.09% | 1.31% | |
| ATC-CNN | 2022 | Chen-2012 | 3883 | S | CNN | 93.01% | 90.72% | 90.53% | 87.77% | 1.53% | |||
| ATC-CNN | 2022 | ATC-SMILES | 4545 | S | CNN | 95.83% | 94.14% | 93.99% | 91.77% | 0.94% | |||
| Method . | Year . | Dataset . | #Drugs . | Rep. . | Model . | Aiming . | Coverage . | Accuracy . | Absolute . | Absolute . | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| . | . | . | . | . | . | . | . | . | |$\uparrow $| . | |$\uparrow $| . | |$\uparrow $| . | True |$\uparrow $| . | False |$\downarrow $| . |
| Chen et al. [4] | 2012 | Chen-2012 | 3883 | I | S | Similarity Search | 50.76% | 75.79% | 49.38% | 13.83% | 8.83% | ||
| iATC-mISF[5] | 2017 | Chen-2012 | 3883 | I | S | F | ML-GKR | 67.83% | 67.10% | 66.41% | 60.98% | 5.85% | |
| iATC-mHyb[6] | 2017 | Chen-2012 | 3883 | I | S | F | O | ML-GKR | 71.91% | 71.46% | 71.32% | 66.75% | 2.43% |
| EnsLIFT[7] | 2017 | Chen-2012 | 3883 | I | S | F | LIFT | 78.18% | 75.77% | 71.21% | 63.30% | 2.85% | |
| EnsANet_LR[9] | 2018 | Chen-2012 | 3883 | I | S | F | CNN,LIFT,RR | 75.40% | 82.49% | 75.12% | 66.68% | 2.62% | |
| EnsANet_LR|$\otimes $|DO[9] | 2018 | Chen-2012 | 3883 | I | S | F | O | CNN,LIFT,RR | 79.57% | 83.35% | 77.78% | 70.90% | 2.40% |
| ATC-NLSP[10] | 2019 | Chen-2012 | 3883 | I | S | F | NLSP | 81.35% | 79.50% | 78.28% | 74.97% | 3.43% | |
| iATC-NRAKEL[11] | 2020 | Chen-2012 | 3883 | I | S | RAKEL,SVM | 78.88% | 79.36% | 77.86% | 75.93% | 3.63% | ||
| iATC-FRAKEL[12] | 2020 | Chen-2012 | 3883 | F | RAKEL,SVM | 78.51% | 78.40% | 77.21% | 75.11% | 3.70% | |||
| FUS3[14] | 2020 | Chen-2012 | 3883 | I | S | F | CNN,LSTM,LIFT,RR | 87.55% | 69.73% | 73.46% | 68.71% | 2.38% | |
| FUS3|$\otimes $|DO[14] | 2020 | Chen-2012 | 3883 | I | S | F | O | CNN,LSTM,LIFT,RR | 79.79% | 84.22% | 79.64% | 73.04% | 2.09% |
| iATC_Deep-mISF[13] | 2020 | Chen-2012 | 3883 | I | S | F | O | DNN | 74.70% | 73.91% | 71.57% | 67.01% | 0.00% |
| CGATCPred[15] | 2021 | Chen-2012 | 3883 | I | S | E | A | CNN,GCN | 81.94% | 82.88% | 80.81% | 76.58% | 2.75% |
| EnsATC[18] | 2022 | Chen-2012 | 3883 | I | S | F | hMuLab,LSTM | 91.39% | 84.32% | 83.38% | 80.09% | 1.31% | |
| ATC-CNN | 2022 | Chen-2012 | 3883 | S | CNN | 93.01% | 90.72% | 90.53% | 87.77% | 1.53% | |||
| ATC-CNN | 2022 | ATC-SMILES | 4545 | S | CNN | 95.83% | 94.14% | 93.99% | 91.77% | 0.94% | |||
Representation (Rep.) abbreviations: I - Chemical interactions, S - Chemical structural features, F - Molecular fingerprint features, O - Drug ontology information, E - Pre-trained word embedding, and A - ATC codes association information.
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
