Drug discovery

Drug discovery

Our approach to drug discovery relies on both applications and developments of new methodology. This lab has developed two important drug discovery tools.

CombiSMoG: Combinatorial Small Molecule Growth

CombiSMoG “grows” molecules directly in the active site of a protein. The growth starts by docking a simple organic molecule into the active site of the target protein. New chemical groups are then added to the small molecule in order to increase the binding affinity of the small molecule to protein. A knowledge-based potential developed to describe protein-small molecule interactions is used to calculate binding affinities, and a Monte Carlo approach is used to decide whether to accept of reject a new chemical group. Efficacy of CombiSMoG has been demonstrated through the design of novel recorded binding inhibitors for the enzyme Carbonic Anhydrase (see this article).

FOG: Fragment Optimized Growth

The approach of FOG is different. The algorithm is first trained on a database made of drugs or natural products, and then it extracts trends and patterns from this database. This information is then used to grow new molecules, which will "look like" the molecules in the initial database. Applied to drug discovery, this allows us to discover new compounds that are more stable or water-soluble than we would on average with random growth. Screening tools were also developed to increase the similarity between the initial database and the new molecules. The FOG-generated structures can then be docked in the active site of a protein.

Next-Generation Algorithms

We are currently working on development of next generation theory and algorithms. The aim is to design better (more drug-like) molecules, to predict their affinities with greater accuracy, to take into account protein flexibility and the role of water in the binding, and to enhance binding specificity in order to reduce side-effect binding and toxicity. All these algorithms are applied to current problems, such as the discovery of new inhibitors to fight against bacteria, viruses, cancer or Alzheimer disease...

For more information, check out what Nicolas ChéronQington Zhou and Michael Zimet are working on.

Selected papers on drug discovery:

  1. De novo design: balancing novelty and confined chemical space
  2. FOG: Fragment Optimized Growth Algorithm for the de Novo Generation of Molecules occupying Druglike Chemical Space
  3. SMoG: de Novo Design Method Based on Simple, Fast, and Accurate Free Energy Estimates. 1. Methodology and Supporting Evidence
  4. SMoG: de Novo Design Method Based on Simple, Fast, and Accurate Free Energy Estimates. 2. Case Studies in Molecular Design