Enormous free energy of binding

[farmaceut]

Hello,

I followed this tutorial to dock my new ligand into another protein. However, in the end, I obtained extremely high affinity values:

mode |   affinity | dist from best mode
     | (kcal/mol) | rmsd l.b.| rmsd u.b.
-----+------------+----------+----------
   1    1.107e+07          0          0
   2    1.115e+07       1.84      3.158
   3    1.155e+07      1.568      6.297
   4    1.833e+07      1.961       2.21
   5    1.987e+07       2.17      3.457
   6    2.265e+07      1.591      2.251
   7    2.525e+07      1.821      2.733
   8    2.624e+07      2.066      3.882
   9     3.51e+07       2.49      4.033

Here are the steps I followed:

1. I downloaded the 4lxd receptor from the PDB, removed everything except the protein, protonated it using Propka as implemented on the PDB2PQR webpage; then manually converted the resulting PQR format to PDB, and ran:
   prepare_receptor -r 4lxd.pdb -o 4lxd.pdbqt

4lxd.pdb
4lxd.pdbqt

2. Then, I prepared the ligand (H4Fst) using:
   mk_prepare_ligand.py -i H4Fst.mol -o H4Fst.pdbqt

H4Fst.mol
H4Fst.pdbqt

3. I generated affinity maps for the AutoDock force field using:
   pythonsh ./prepare_gpf.py -l H4Fst.pdbqt -r 4lxd.pdbqt -y
   and obtained the 4lxd.gpf file, which I modified using Chimera visualization:

I set npts 22 12 11 and gridcenter 21.3455 31.4093 8.11142 leaving the remaining parameters as they were.
4lxd.gpf

4. I ran AutoGrid4 using:
   autogrid4 -p 4lxd.gpf -l 4lxd.glg
   4lxd.glg

and obtained all the maps enclosed in this file:
maps.zip

5. Finally, I started Vina using the command:
   vina --ligand H4Fst.pdbqt --maps 4lxd --scoring ad4 --exhaustiveness 32 --out output.pdbqt
   and the output can be found here:
   output.pdbqt.txt

Appreciate advice where I made the issue.
Full_Archive.zip

[rwxayheee]

Hi @farmaceut
In PyMOL, the dimension of the box is displayed in Å
npts 22 12 11 will generate a small box of
0.375 * 22 = 8.25 Å
0.375 * 12 = 4.5 Å
0.375 * 11 = 4.125 Å
where 0.375 Å is the spacing of the grid

[farmaceut]

Hi again @rwxayheee :) Thanks for quick response. So the solution would be either shifting spacing to 1 or change enpts following the equations

0.375 * X = 22
0.375 * Y = 12
0.375 * Z = 11

so npts 60 32 30 in that case?

[rwxayheee]

Hi @farmaceut Yes npts 60 32 30 will give you the defined box in PyMOL without changing the spacing :>

[farmaceut]

@rwxayheee
Albeit, 4lxd.gpf looks now like that:

npts 60 32 30                        # num.grid points in xyz
gridfld 4lxd.maps.fld                # grid_data_file
spacing 0.375                        # spacing(A)
receptor_types A C NA OA N SA HD     # receptor atom types
ligand_types A HD OA                 # ligand atom types
receptor 4lxd.pdbqt                  # macromolecule
gridcenter 21.3455 31.4093 8.11142   # xyz-coordinates or auto
smooth 0.5                           # store minimum energy w/in rad(A)
map 4lxd.A.map                       # atom-specific affinity map
map 4lxd.HD.map                      # atom-specific affinity map
map 4lxd.OA.map                      # atom-specific affinity map
elecmap 4lxd.e.map                   # electrostatic potential map
dsolvmap 4lxd.d.map              # desolvation potential map
dielectric -0.1465                   # <0, AD4 distance-dep.diel;>0, constant

docking yielded same incorrect values:

mode |   affinity | dist from best mode
     | (kcal/mol) | rmsd l.b.| rmsd u.b.
-----+------------+----------+----------
   1    1.048e+07          0          0
   2    1.055e+07       2.02      3.311
   3    1.073e+07      1.719      6.845
   4    1.097e+07      1.944      2.312
   5      1.1e+07      1.262      6.415
   6    1.112e+07      2.075      6.526
   7    1.123e+07       2.08      6.958
   8    1.203e+07      1.709      2.266
   9    1.208e+07      1.882       2.61

[rwxayheee]

@farmaceut
I got something like

mode |   affinity | dist from best mode
     | (kcal/mol) | rmsd l.b.| rmsd u.b.
-----+------------+----------+----------
   1       -4.222          0          0

Did you run autogrid4 with the new gpf file?

here's my output:
output.pdbqt.txt

[farmaceut]

@rwxayheee
My issue! Forgot to run autogrid4 . Now it is more than good. Thank you!
'''
mode | affinity | dist from best mode
| (kcal/mol) | rmsd l.b.| rmsd u.b.
-----+------------+----------+----------
1 -4.208 0 0
2 -4.072 1.765 6.758
3 -3.961 1.832 2.393
4 -3.942 2.095 2.551
5 -3.857 2.16 6.683
6 -3.636 4.113 7.674
7 -3.548 2.621 3.775
8 -3.291 4.181 6.308
9 -3.019 3.248 7.181
'''