Filter outputs of Docking

This document describes how to filter the outputs of Docking based on various properties.

Here we assume that you have constructed a Complex object and successfully run Docking. Following convention, we assume that the Complex object is called sim.

Fetch Docking results

First, we get the results of Docking in a pandas DataFrame using:

df = sim.docking.get_results()

Fetch combined SDF

We can generate a single SDF file with all the poses using:

sim.docking.get_poses("poses.sdf")

Filter results

We can now filter this dataframe using any criteria we want. For example, we can only retain ligands that have a pose_score greater than 0.9 using:

df = df[df["pose_score"] > 0.9]

The filtered dataframe now only has ligands that matches the required criterion.

Filter Ligands by SMILES strings

We can create a new SDF file with only these ligands using:

from deeporigin.drug_discovery import Ligand

ligands = Ligand.from_sdf("..path/to/poses.sdf") # from docking 

smiles_strings = df["SMILES"]

ligands = [ligand for ligand in ligands if ligand.properties["SMILES"] in smiles_strings]