Getting your Compounds in Shape

& Maria Brandl

We know that most pharmaceutically relevant compounds bind to their targets in a relaxed conformation. The challenge in discovery is to figure out rapidly which conformations are readily accessible for the molecules we are considering. There is now a new solution to address this based on statistical, rather than just energetic approaches.

Driven by the wealth and diversity of bond, angle and torsion information in the Cambridge Structural Database, the CSD Conformer Generator produces realistic ensembles of low energy ligand structures. These are ready to be exploited for drug design in the presence and also in the absence of detailed knowledge about the three-dimensional structure of the protein active site.

Our applications scientists recently challenged our CSD driven Conformer Generator to re-create protein bound ligand conformations.1 Using the CSD Python API, our colleagues Ilenia and Alicia were easily able to create a work flow to generate conformations and check them against ligands from the PDB. It was a success - our conformer generator was able to create closely matched ligand conformations in more than 95% of cases!  

The image below shows one of our generated conformers (blue) together with the experimental ligand structure (magenta), bound to the anti-TB target –methionine aminopeptidase (PDB:3pka). Intriguingly, the ligand generated by the CSD Conformer Generator is in a perfect shape to form its key contacts to protein side chains and Mn­2+ ions.


The CSD Conformer Generator has just been released within the CSD-Materials, CSD-Discovery and CSD-Enterprise packages. Conformers can be generated from molecules loaded into Mercury, the CSD’s interface for viewing and analysing small organic molecule crystal structures. 

We invite you to evaluate the performance of CSD Conformer Generator on your own choice of PDB-structures. Just collect a list of PDB-codes and use Alicia’s ‘’ program posted on the CSD Python API Forum.

The CSD Conformer Generator’s functionality is also fully available via the CSD Python API, allowing you seamlessly to link conformer generation to other applications and enabling complex workflows and analyses.

To create the conformers (Y02_conformers.sdf) for the anti-tuberculosis agent Y02 (Y02.mol), via the CSD Python API, all you need are a few very simple python statements.

from ccdc import io
from ccdc import conformer
mol_reader = io.MoleculeReader('Y02.mol')
mol = mol_reader[0]
conformer_generator = conformer.ConformerGenerator()
conformer_generator.settings.superimpose_conformers_onto_reference = True
conformers = conformer_generator.generate(mol)
with io.MoleculeWriter('Y02_conformers.sdf') as mol_writer:
for c in conformers:

We hope the CSD Conformer Generator proves to be a useful addition to your discovery tool set!

  1. Giangreco, et al. "Assessment of a Cambridge Structural Database-Driven Overlay Program."Journal of Chemical Information and Modeling 11 (2014): 3091-3098