Ligands of Howard et al, JMC 2006, 49, 1346

For the ligands just IC50 values were given. These were not used in the training sets, but the ligands could be used as an external test set.

 

local_start.sh -sdf /home/henricsn/combine2go/sdf/all_071106_3D.sdf \

-inrec ../model3_180506/ -inlig ./ -model thrombin-model3.min.pdb \

-ligbase .ligand.pdb |tee local.log

 

calc_inter_model.sh -ki thrombin -x -indir ./ -outdir ./ -sdf \

/home/henricsn/combine2go/sdf/all_071106_3D.sdf -res2id residue2list|tee calc.log

 

 

Data file : /home/henricsn/combine2go/data/thrombin/model3_180506/anal2/output-A12_1-A09-A84-A88-A89-A90.dat

no variable selection, LV4

 

External PLS predictions for /home/henricsn/combine2go/data/thrombin/ligands_Howard/ligands_Howard.dat

 

Results

For the recently published crystal structures of thrombin, which were in complex with a series of ligands binding exclusively to the S1 and/or S2/S3/S4 pockets, IC50 but no Ki values were given. Because of the missing Ki values they were not used as a training set, but under the assumption of similar order of magnitude for IC50 and Ki values the ligands were suitable as an external test set to evaluate the COMBINE model. The predicted ∆G values for the X‑ray and docked conformation of the ligands B29/2c8w, B28/2c8x and id403 (no X-ray structure available) were similar to the experimental ones demonstrating a good quality of the COMBINE model of thrombin. But for ligands, which weakly bound in the experiment either only to the S1 pocket (B26/2c8z, B24/2c90) or left the S1 pocket vacant (B25/2c8y, B27/2c93) (Howard, et al., 2006), could not be predicted correctly as weak binders.


thrombin-B29_blue-0.2_red0.2.png
xray
1   B29.0    id 404               2c8w

Exper: +8.889e+05 Pred:( 1 ) -11.95    ( 2 ) -14.08    ( 3 ) -13.98    ( 4 ) -14.19    ( 5 ) -14.45   

pred Ki = 3.96123e-05 = 0.00004 uM

exper IC50 = 0.0037 uM

exper delta G in kcal/mol= -11.5021

 

docking
best dG bind elec

69  id404.thr.7.9                                     

Exper: +8.889e+05 Pred:( 1 ) -8.795    ( 2 ) -11.24    ( 3 ) -11.66    ( 4 ) -12.28    ( 5 ) -12.33   

wrong docking pose

 

61  id404.thr.7.1                                      

Exper: +8.889e+05 Pred:( 1 ) -8        ( 2 ) -9.639    ( 3 ) -9.695    ( 4 ) -10.46    ( 5 ) -10.64   

better docking pose

 



thrombin-B28_blue-0.2_red0.2.png
xray
2   B28.0    id 402     2c8x                                     

Exper: +8.889e+05 Pred:( 1 ) -11.7     ( 2 ) -13.76    ( 3 ) -13.49    ( 4 ) -13.73    ( 5 ) -14.03   

pred Ki = 8.61072e-05=0.00009 uM

exper IC50 = 0.22 uM

exper delta G in kcal/mol= -9.08186

 

docking
best dG bind elec

46  id402.thr.5.6                                      

Exper: +8.889e+05 Pred:( 1 ) -9.396    ( 2 ) -11.68    ( 3 ) -12.37    ( 4 ) -12.62    ( 5 ) -12.37   

wrong docking pose

 

43  id402.thr.5.3                                     

Exper: +8.889e+05 Pred:( 1 ) -8.858    ( 2 ) -10.95    ( 3 ) -10.69    ( 4 ) -10.29    ( 5 ) -10.07   

better docking pose

 


thrombin-B25_blue-0.2_red0.2.png

xray
3   B25.0  id 399                 2c8y                           

Exper: +8.889e+05 Pred:( 1 ) -9.896    ( 2 ) -11.75    ( 3 ) -11.41    ( 4 ) -11.52    ( 5 ) -11.41   

pred Ki = 0.00359012

exper IC50 = 100 uM

exper delta G in kcal/mol= -5.45655

 

docking
best dG bind elec

19  id399.thr.2.9                                     

Exper: +8.889e+05 Pred:( 1 ) -8.417    ( 2 ) -10.25    ( 3 ) -10.26    ( 4 ) -11.01    ( 5 ) -11.04   

wrong docking pose

 

The bad prediction is propably based on the vacant S1 pocket.

 

docking pose in high energetic open conformation but in Xray a hydrophobic collapse was observed

fragment

thrombin-B26_blue-0.2_red0.2.png
xray
4   B26.0  id 400     2c8z                                       

Exper: +8.889e+05 Pred:( 1 ) -7.298    ( 2 ) -8.516    ( 3 ) -7.716    ( 4 ) -8.242    ( 5 ) -8.203   

pred Ki = 0.908012

exper IC50 > 1000 uM

exper delta G in kcal/mol < -4.09

 

docking
best dG bind elec

26  id400.thr.3.6                                     

Exper: +8.889e+05 Pred:( 1 ) -6.372    ( 2 ) -6.114    ( 3 ) -6.124    ( 4 ) -6.001    ( 5 ) -5.823   

wrong docking pose

 

fragment

thrombin-B24_blue-0.2_red0.2.png
xray
5   B24.0  id 398              2c90                              

Exper: +8.889e+05 Pred:( 1 ) -7.826    ( 2 ) -10.19    ( 3 ) -9.976    ( 4 ) -10.1     ( 5 ) -9.992   

pred Ki = 0.0394511

exper IC50 = 330 uM

exper delta G in kcal/mol= -4.74923

 

docking
best dG bind elec

4   id398.thr.1.4                                     

Exper: +8.889e+05 Pred:( 1 ) -7.944    ( 2 ) -9.953    ( 3 ) -9.815    ( 4 ) -9.727    ( 5 ) -9.565   

wrong docking pose

 

The prediction for this small fragment is bad, but e. g. benzamidine with similar size were not used in the training set, too, becaue of beeing an outlier or an inconsistancy in the Ki values.



thrombin-B27_blue-0.2_red0.2.png
xray
6   B27.0   id 401   2c93                                       

Exper: +8.889e+05 Pred:( 1 ) -10.12    ( 2 ) -11.9     ( 3 ) -11.4     ( 4 ) -11.32    ( 5 ) -11.24   

pred Ki = 0.00503178

exper IC50 = 12 uM

exper delta G in kcal/mol= -6.71268

 

docking
best dG bind elec

38  id401.thr.4.8                                     

Exper: +8.889e+05 Pred:( 1 ) -8.167    ( 2 ) -10.39    ( 3 ) -10.5     ( 4 ) -11.14    ( 5 ) -11.34   

wrong docking pose

 

Bad prediction due to a vacant S1 pocket (similar to B25).

no xray available
id 403

exper IC50 = 0.0014 uM

exper delta G in kcal/mol= -12.0779

 

docking
51  id403.thr.6.1                                     

Exper: +8.889e+05 Pred:( 1 ) -8.238    ( 2 ) -10.11    ( 3 ) -10.27    ( 4 ) -10.42    ( 5 ) -10.31   

wrong docking pose

 

53  id403.thr.6.3                                     

Exper: +8.889e+05 Pred:( 1 ) -9.193    ( 2 ) -11.22    ( 3 ) -11.18    ( 4 ) -11.52    ( 5 ) -11.65   

best docking pose

Privacy Imprint