see the app
see intro post for motivation.
This is a streamlit app to visualize docking hits for the purposes of manual curation before moving to in vitro testing. You might use this to prioritize a ranked list of docking hits so that you only spend money on buying the ligands with the best chance of successfully binding the target.
In this test case, the 5 example ligands have been docked against the D4 receptor, crystal structure 5WIU bound to nemonapride. The ligands were pulled from a public dataset in doi). Smina was used for docking, and obabel for file conversion to/from pdbqt/pdb. The docking can be done within run_smina.ipynb assuming you have smina installed.
Load the app, and select whether or not to use Annotations that describe the pharmacophores (these can be helpful but sometimes messy). Write a number, from 0 - 4 inclusive, and press enter to load one of the docked ligand poses. How you judge the likelihood of a hit is up to you ;)
Have a set of docked ligands named 0.pdb, 1.pdb, etc... in the files dir, up to as many as you want. You'll also need to have a receptor pdb file in the files dir, and change the name in stApp.py to the name of your receptor. Same goes for the co-crystallized ligand.
This pulls together some amazing libraries written by others:
py3dmolfor 3d visualization: https://3dmol.csb.pitt.edu/ , Rego & Koes: https://doi.org/10.1093/bioinformatics/btu829- molecular properties are calculated with the
RDKit: https://www.rdkit.org/ - docking used
smina: https://sourceforge.net/projects/smina/ - docking pipeline was described by
Esben Jannik Bjerrumat https://www.cheminformania.com/ligand-docking-with-smina/ - ligands are from a publicly available dataset from
Lyu et al: https://doi.org/10.1038/s41586-019-0917-9 - structure file conversions by obabel: https://openbabel.org/docs/dev/Command-line_tools/babel.html
- and the pieces were put together by lewis martin: https://ljmartin.github.io/