Table 2 First four rows: example SMILES strings from the indication-to-drug task; Last four rows: example MolT5 indication generations from the drug-to-indication task.
Input | Ground truth | Output | Validity/similarity |
|---|---|---|---|
Indication-to-drug | |||
Diabetes mellitus | COCCOc1cnc(NS(=O)(=O)c2ccccc2)nc1 | O=C([O-])CC(=O)[O-] | Valid |
Coronary artery disease | CCOC(=O)C(C)=O | CCCCC[C@H](O)CC=CCC=CCCCC(=O)O | Valid |
Respiratory system disease | CCC1(C)CC(=O)NC(=O)C1 | [H+].C(=O)[O-])[O-] | Syntax Error |
Hemorrhage | CC1=CC(=O)c2ccccc2C1=O | C(=O)C(=O)O)O.O)O.O)O.O.O | Syntax Error |
Drug-to-indication | |||
CN(C)CCOC(c1ccccc1)c1ccccc1 | Allergic disease ... cancer ... eczema ... | ... and cancer ... | 0.2206 |
CCc1cc(C(N)=S)ccn1 | Multidrug-resistant tuberculosis osteomyelitis ... | ... cancer | 0.2262 |
O=C([O-])c1ccccc1.[Na+] | Encephalopathy psychosis | Inamideamide protein protein proteinamide. | 0.0183 |
Clc1ccccc1CN1CCc2sccc2C1 | Internal carotid artery stenosis ... Recurrent thrombophlebitis | Amideamideamide. | 0.0316 |
