Notebook to look at the docking results

This uses the results from the SMINA docking run using the ConformationGenerationDocking.ipynb

In [7]:
import sys
from collections import defaultdict
import numpy as np
import py3Dmol
from rdkit import Chem
from rdkit.Chem import AllChem
from rdkit.Chem.Draw import IPythonConsole
from rdkit.Chem import PandasTools
import pandas as pd
IPythonConsole.ipython_3d=True
%pylab inline

Populating the interactive namespace from numpy and matplotlib
In [8]:
# File locations, this is the final output from ConformationGenerationDocking.ipynb
DockingsdfFilePath = 'Alldata.sdf.gz' # The results of the docking experiment.

Import the docking/scoring results into a dataframe

In [9]:
docked_df = PandasTools.LoadSDF(DockingsdfFilePath,molColName='Molecule', removeHs=False)
In [10]:
docked_df.head(5) #view the first 5 rows
Out[10]:
minimizedAffinity name RFScoreVS_v2 ID Molecule
0 -5.30895 ART 15342099 6.148551 ART 15342099 Mol
1 -5.26728 ART 15342099 6.127652 ART 15342099 Mol
2 -5.04994 ART 15342099 6.090561 ART 15342099 Mol
3 -4.74888 ART 15342099 6.067428 ART 15342099 Mol
4 -4.72620 ART 15342099 6.106997 ART 15342099 Mol

Sort the rows on either minimizedAffinity or the scoring funtion RFScoreVS_v2

In [11]:
#docked_df.sort_values(["minimizedAffinity"], axis=0, ascending=False, inplace=True) #sort by minimised affinity
docked_df.sort_values(["RFScoreVS_v2"], axis=0, ascending=False, inplace=True) #or sort by scoring function
In [12]:
docked_df.head(5) # rows should now be sorted with best scoring top
Out[12]:
minimizedAffinity name RFScoreVS_v2 ID Molecule
4632 -5.03851 SYN 13975176 6.312364 SYN 13975176 Mol
185 -3.77326 ASN 10790639 6.305832 ASN 10790639 Mol
356 -5.91577 AEM 10028511 6.30018 AEM 10028511 Mol
4565 -5.96147 SYN 13975163 6.298634 SYN 13975163 Mol
4059 -5.11896 LMG 16204894 6.294883 LMG 16204894 Mol

Export the top scoring molecules, edit "head(1)" to the number you want to export

In [13]:
selectedPose = 'selectedpose.sdf'
In [14]:
PandasTools.WriteSDF(docked_df.head(5), selectedPose, molColName="Molecule", idName="ID", properties=list(docked_df.columns))
In [15]:
selecteddocked_df = PandasTools.LoadSDF(selectedPose,molColName='Molecule', removeHs=False)
selecteddocked_df
Out[15]:
minimizedAffinity name RFScoreVS_v2 ID Molecule
0 -5.03851 SYN 13975176 6.312364 SYN 13975176 Mol
1 -3.77326 ASN 10790639 6.305832 ASN 10790639 Mol
2 -5.91577 AEM 10028511 6.30018 AEM 10028511 Mol
3 -5.96147 SYN 13975163 6.298634 SYN 13975163 Mol
4 -5.11896 LMG 16204894 6.294883 LMG 16204894 Mol

Adding hydrogens to the molecules using Openbabel (conda install openbabel)

In [16]:
m = Chem.MolFromMolFile(selectedPose, removeHs=False) #View first structure, No hydrogens
m

You appear to be running in JupyterLab (or JavaScript failed to load for some other reason). You need to install the 3dmol extension:
jupyter labextension install jupyterlab_3dmol

Out[16]:

Using pybel - A Python module that simplifies access to the Open Babel API

In [17]:
selectedPose = 'selectedpose.sdf'
selectedPoseH = 'selectedposeH.sdf' #added hydrogens
from openbabel import pybel

largeSDfile = pybel.Outputfile("sdf", selectedPoseH, overwrite=True)

for molecule in pybel.readfile("sdf", selectedPose):
    molecule.OBMol.AddHydrogens()
    largeSDfile.write(molecule)
In [ ]:
In [18]:
mH = Chem.MolFromMolFile(selectedPoseH, removeHs=False) #View first structure, Hydrogens present
mH

You appear to be running in JupyterLab (or JavaScript failed to load for some other reason). You need to install the 3dmol extension:
jupyter labextension install jupyterlab_3dmol

Out[18]:

The docked structures can now be opened with PYMOL together with the protein structure (protein_minus_ligand.pdb).

In [31]:
#first type pymol R in a Terminal window to start pymol
In [38]:
import xmlrpc.client as xmlrpclib
In [39]:
ProteinForDocking = 'protein_minus_ligand.pdb'
selectedPoseH = 'selectedposeH.sdf' #
In [40]:
srv = xmlrpclib.ServerProxy('http://localhost:9123')
In [41]:
svrcommand = 'load ' + ProteinForDocking
srv.do(svrcommand)
In [42]:
svrcommand = 'load ' + selectedPoseH
srv.do(svrcommand)
In [43]:
srv.do ('select prot, (protein)')
srv.do ('show surface, prot')
In [ ]:
In [ ]: