Import libraries


In [1]:
from base64 import b64encode
import numpy as np
import pandas as pd
from IPython.display import HTML

import oddt
import oddt.pandas as opd
from oddt.fingerprints import PLEC

Protein-Ligand Extended Connectivity Fingerprint (PLEC)

This notebook provides a short guide on how to use Protein-Ligand Extended Connectivity (PLEC) Fingerprint for a series of docked ligands. The output is in CSV format which should be easy to use with other, external software packages.

More informatioFor more information about PLEC see publication.

Load receptor (protein) file in pdb, mol2 or other formats


In [2]:
receptor = next(oddt.toolkit.readfile('pdb', 'xiap/receptor_rdkit.pdb'))

Load ligands file (docked conformations)


In [3]:
data = opd.read_sdf('xiap/actives_docked.sdf')

Set PLEC FP parameters:

size: the lenght of final fingerprint (default=16384)

depth_protein: the environment diameter on protein side in bond steps (default=4)

depth_ligand: the environment diameter on ligand side in bond steps (default=2)

distance_cutoff: the distance between two atoms to be considered interacting (derfault=4.5)

Generate PLEC FPs


In [4]:
data['plec'] = data['mol'].map(lambda x: list(PLEC(x, protein=receptor, 
                                                   size=4092, 
                                                   depth_protein=5,
                                                   depth_ligand=1,
                                                   distance_cutoff=4.5,
                                                   sparse=False
                                                  )))

In [5]:
data[['mol', 'name', 'act', 'plec']].head()


Out[5]:
mol name act plec
0 312335 6 [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 4, 1, 0, 0, 1, 0, 0, 0, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0, 2, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, ...]
1 312335 6 [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 3, 1, 0, 0, 1, 0, 0, 0, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5, 1, 0, 0, 0, 2, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, ...]
2 312335 6 [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 1, 0, 0, 2, 0, 0, 0, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5, 1, 0, 0, 0, 2, 0, 0, 0, 3, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, ...]
3 312370 10 [0, 0, 0, 0, 0, 0, 0, 1, 5, 0, 0, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 3, 4, 0, 0, 0, 1, 0, 0, 0, 9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 6, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, ...]
4 312335 6 [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 4, 1, 0, 0, 2, 0, 0, 0, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5, 1, 0, 0, 0, 2, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, ...]

In [ ]:

Download CSV with PLEC FP


In [6]:
HTML('<a href="data:application/binary;base64,%s" download="fingerprint.csv">Download CSV</a>' % b64encode(data.to_csv().encode()).decode())


Out[6]: