Skip to main content
. Author manuscript; available in PMC: 2011 Jun 6.
Published in final edited form as: J Comput Aided Mol Des. 2008 Feb 14;22(5):269–286. doi: 10.1007/s10822-008-9174-y

Table 2. Proteins and known actives selected as training data.

The eight protein targets used to test the scoring function optimization approach are the overlap between the Pham set [13] and the DUD set [2]. The known ligands used for training (Nmols) are from the Pham set. Also shown are the mean physical properties for each active set. Surflex treats polar atoms as donors or acceptors if not formally charged and scales the functional term if the atoms are formally charged. The mean number of positive and negative polar atoms is shown, with the mean number of formally charged atoms in parentheses.

Target PDB
code
Nmols
Train
MW Nrot N Polar Negative
(formally
charged)
N Polar Positive
(formally
charged)

AChE 1e66 6 334.7 4.8 1.3 (0.0) 4.2 (2.7)
ER 3ert 10 465.6 9.9 3.3 (0.0) 2.6 (0.8)
FXa 1fjs 6 450.7 8.3 4.8 (0.7) 4.7 (3.2)
HIVPR 1pro 20 559.7 14.1 5.7 (0.3) 4.8 (1.0)
PARP 2pax 15 202.1 0.7 2.4 (0.0) 1.3 (0.0)
Thrombin 1c4v 20 429.2 8.8 3.6 (0.3) 5.2 (3.6)
TK 1kim 10 264.5 4.9 4.7 (0.0) 3.8 (0.0)
Trypsin 1qbo 20 386.8 6.5 3.5 (0.6) 5.6 (4.6)