Name | Equations | Parameters |
---|
MIN-SIM
| 18,1 | 0 |
MAX-SIM
| 17,1 | 0 |
SUM-SIM
| 16,1 | 0 |
NUMDEN-SIM
| 23,19,1 | 0 |
MIN-RANK
| 15,1 | 0 |
MAX-RANK
| 14,1 | 0 |
SUM-RANK
| 13,1 | 0 |
BAYES†| 24 | 0 |
SUM-EH*†| 16,11 | 2 |
SUM-ET* | 16,12 | 2 |
SUM-TP* | 16,9 | 1 |
BKD†| 20,11 | 2 |
ETD* | 20,12 | 2 |
TPD* | 20,9 | 1 |
- A list of abbreviated names for different multiple molecule queries. In the case of any methods with weight vectors, we use equally weighted convex combinations, in other words, wi = 1/|| where || is the number of query molecules. So, for example, the SUM-SIM becomes the average similarity. Methods introduced in this study for the first time are marked with a star (*). Methods using the compressed metrics are marked with a dagger (†); the rest use corrected metrics.