Skip to main content

Table 5 Cases where the optimist query outperforms the opportunist one in CPU-bounded experiments

From: Consensus queries in ligand-based virtual screening experiments

Size

Speedup

MLQSAR

NRLIST

5

1.46

6/9

9/10

10

2.05

7/9

8/10

20

3.19

7/9

8/10

  1. Size: consensus size; speedup: how many times the optimist consensus is faster at scoring than the opportunist consensus. Experiment: compute the accumulated curve of actives on MLQSAR (median curve over 10 experiments) and NRLIST (median curve over 50 experiments). The number of times where the optimist curve dominates the opportunist one is reported