ADL: Best way to assign partial charges

Douglas Houston dhouston at staffmail.ed.ac.uk
Wed Apr 29 08:45:19 PDT 2009


The short answer appears to be that it makes a small difference.

I tested by docking 9000 compounds that had been assigned partial charges
using the ADT python script. I then repeated this to get an idea of the
"dispersion" or reproducibility of the docking. The Spearman's rank
correlation coefficient for the 9000 compounds ranked by Estimated Free
Energy of Binding between the two runs was 0.8 (so not too much dispersion).

I then docked the same 9000 compounds but this time with partial charges
assigned by OpenBabel. Everything else was kept the same. The Spearman's
coefficient between this run's ranking and the other 2 runs was 0.7 in both
cases. 

Therefore there was a minor difference in how Autodock ranked the compounds,
but nothing really to worry about.

(This test isn't exhaustive but it takes a while to dock 9000 compounds!)



许文 wrote:
> 
> Hi,
> that's an intresting problem maybe you can tell everyone is that charge
> difference make the docking results different after your experimental
> docking runs.
> thank for sharing your experience.
> 
> 
> 
> 
> 在2009-04-21,"Douglas Houston" <dhouston at staffmail.ed.ac.uk> 写道:
>>
>>It seems that the Gasteiger charges calculated by prepare_ligand4.py are
>>indeed different from the Gasteiger charges calculated by OpenBabel.
>>
>>Most of them aren't that different, but some are significantly different.
>>For example, atom 5 in this molecule below is quite negatively charged
using
>>one method, and slightly positively charged using the other. Since
nitrogen
>>is more electronegative, it seems to me that it would draw electrons from
>>the carbons it's bonded to and would indeed have a small negative charge,
so
>>it appears that, in this case, the OpenBabel method is more "correct"
>>(please correct me if I'm wrong).
>>
>>Charges added by OpenBabel 2.2.1:
>>
>>ROOT
>>ATOM      1  C   LIG d  11       0.000   0.000   0.000  0.00  0.00   
-0.002
>>C 
>>ATOM      2  C   LIG d  11      -0.595  -1.258  -0.587  0.00  0.00   
-0.034
>>A 
>>ATOM      3  C   LIG d  11       0.699  -1.260   2.083  0.00  0.00    
0.065
>>C 
>>ATOM      4  C   LIG d  11       0.000   0.000   1.540  0.00  0.00    
0.014
>>C 
>>ATOM      5  N   LIG d  11       2.106  -1.152   1.669  0.00  0.00   
-0.295
>>N 
>>ATOM      6  C   LIG d  11      -0.599  -2.453   0.142  0.00  0.00   
-0.042
>>A 
>>ATOM      7  C   LIG d  11      -0.007  -2.512   1.530  0.00  0.00    
0.054
>>C 
>>ATOM      8  C   LIG d  11       1.476   0.000  -0.440  0.00  0.00    
0.027
>>C 
>>ATOM      9  C   LIG d  11       2.245  -1.168   0.206  0.00  0.00    
0.086
>>C 
>>ATOM     10  C   LIG d  11      -1.148  -1.220  -1.872  0.00  0.00    
0.049
>>A 
>>ATOM     11  C   LIG d  11      -1.154  -3.611  -0.415  0.00  0.00    
0.007
>>A 
>>ATOM     12  C   LIG d  11      -1.704  -2.378  -2.429  0.00  0.00    
0.117
>>A 
>>ATOM     13  C   LIG d  11      -0.767   1.212  -0.522  0.00  0.00    
0.016
>>C 
>>ATOM     14  C   LIG d  11      -1.707  -3.573  -1.701  0.00  0.00    
0.045
>>A 
>>ATOM     15  C   LIG d  11      -1.433   0.067   2.061  0.00  0.00    
0.010
>>C 
>>ENDROOT
>>BRANCH   5  16
>>ATOM     16  C   LIG d  11       2.873  -2.267   2.244  0.00  0.00    
0.112
>>C 
>>BRANCH  16  17
>>ATOM     17  C   LIG d  11       3.946  -1.722   3.155  0.00  0.00   
-0.013
>>C 
>>ATOM     18  C   LIG d  11       4.773  -2.562   3.793  0.00  0.00   
-0.078
>>C 
>>ATOM     19  C   LIG d  11       5.847  -2.017   4.704  0.00  0.00    
0.038
>>C 
>>ATOM     20  C   LIG d  11       4.631  -4.053   3.600  0.00  0.00    
0.038
>>C 
>>ENDBRANCH  16  17
>>ENDBRANCH   5  16
>>BRANCH  12  21
>>ATOM     21  O   LIG d  11      -2.244  -2.341  -3.688  0.00  0.00   
-0.507
>>OA
>>ATOM     22  H   LIG d  11      -2.242  -1.433  -4.241  0.00  0.00    
0.292
>>HD
>>ENDBRANCH  12  21
>>TORSDOF 3
>>
>>Charges added by prepare_ligand4.py:
>>
>>ROOT
>>HETATM    1  C   LIG     1       0.000   0.000   0.000  1.00  0.00   
-0.009
>>C 
>>HETATM    2  C   LIG     1      -0.595  -1.258  -0.587  1.00  0.00   
-0.040
>>A 
>>HETATM    3  C   LIG     1       0.699  -1.260   2.083  1.00  0.00    
0.232
>>C 
>>HETATM    4  C   LIG     1       0.000   0.000   1.540  1.00  0.00    
0.028
>>C 
>>HETATM    5  N   LIG     1       2.106  -1.152   1.669  1.00  0.00    
0.097
>>N 
>>HETATM    6  C   LIG     1      -0.599  -2.453   0.142  1.00  0.00   
-0.049
>>A 
>>HETATM    7  C   LIG     1      -0.007  -2.512   1.530  1.00  0.00    
0.075
>>C 
>>HETATM    8  C   LIG     1       1.476   0.000  -0.440  1.00  0.00    
0.044
>>C 
>>HETATM    9  C   LIG     1       2.245  -1.168   0.206  1.00  0.00    
0.274
>>C 
>>HETATM   10  C   LIG     1      -1.148  -1.220  -1.872  1.00  0.00    
0.045
>>A 
>>HETATM   11  C   LIG     1      -1.154  -3.611  -0.415  1.00  0.00    
0.010
>>A 
>>HETATM   12  C   LIG     1      -1.704  -2.378  -2.429  1.00  0.00    
0.066
>>A 
>>HETATM   13  C   LIG     1      -0.767   1.212  -0.522  1.00  0.00    
0.021
>>C 
>>HETATM   14  C   LIG     1      -1.707  -3.573  -1.701  1.00  0.00    
0.037
>>A 
>>HETATM   15  C   LIG     1      -1.433   0.067   2.061  1.00  0.00    
0.013
>>C 
>>ENDROOT
>>BRANCH   5  16
>>HETATM   16  C   LIG     1       2.873  -2.267   2.244  1.00  0.00    
0.300
>>C 
>>BRANCH  16  17
>>HETATM   17  C   LIG     1       3.946  -1.722   3.155  1.00  0.00    
0.005
>>C 
>>HETATM   18  C   LIG     1       4.773  -2.562   3.793  1.00  0.00   
-0.090
>>C 
>>HETATM   19  C   LIG     1       5.847  -2.017   4.704  1.00  0.00    
0.042
>>C 
>>HETATM   20  C   LIG     1       4.631  -4.053   3.600  1.00  0.00    
0.042
>>C 
>>ENDBRANCH  16  17
>>ENDBRANCH   5  16
>>BRANCH  12  21
>>HETATM   21  O   LIG     1      -2.244  -2.341  -3.688  1.00  0.00   
-0.361
>>OA
>>HETATM   22  H   LIG     1      -2.242  -1.433  -4.241  1.00  0.00    
0.217
>>HD
>>ENDBRANCH  12  21
>>TORSDOF 3
>>
>>I plan to re-dock a large number of molecules using the new charges and
see
>>if they make a big difference to the rankings.
>>-- 
>>View this message in context:
http://n2.nabble.com/Best-way-to-assign-partial-charges-tp2642998p2669536.html
>>Sent from the AutoDock mailing list archive at Nabble.com.
>>
>>________________________________________________
>>--- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---
> ________________________________________________
> --- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---
> 

-- 
View this message in context: http://n2.nabble.com/Best-way-to-assign-partial-charges-tp2642998p2741378.html
Sent from the AutoDock mailing list archive at Nabble.com.




More information about the autodock mailing list