Structure-based
Virtual Screening
Instead of using a conventional high-throughput screening approach, the industry standard for many years, which requires large amounts of chemical reagents, disposables and solutions, we use in silico screening approaches to hierarchically select small compounds from a chemical space containing more than 750 million purchasable compounds. In the final step of our screening approach, we typically select around the top 1000 compounds for in vitro testing and validation of biological activity. From this first round of focused hit finding we design further routes of hit optimization, including existing and newly synthesized molecules. By using in silico methods, we can significantly reduce the amounts of time investment, chemical waste and test consumables and more importantly we can accelerate the molecular discovery process.
Structure-based
Peptide Design
Peptides are short chains of interconnected amino acid building blocks. To create even the smallest peptide, containing just 2 amino acids, there are at least 400 possibilities (20 natural amino acids for each position). Thus, the development of therapeutic peptides, that typically contain between 5 to 97 amino acids, is an absolutely challenging task. To accelerate the discovery of “lead” peptidic inhibitors/activators, we apply structure-based methods using as much as possible the examples that nature provides us in the form of known structures of complexes between drug targets and their physiological binding partners. These starting structures are then in silico evolved to obtain peptides with optimal binding strength and specificity.
Our approach allows us to significantly reduce the number of peptides that need to be synthesized and experimentally tested.
As an illustration, instead of synthesizing 1,000 or 1,000,000 peptides, we have applied our computer-based methods to create a shortlist of around 10-20 peptide candidates (in silico). From these candidates, we can select 5-10 peptides (the top candidates according to their predicted binding affinity) for synthesis and in vitro tests, and the most active peptides (1-2 peptides) next will be selected for in vivo testing.
Protein Engineering
Similar to peptide design, we have utilized the structure-based approaches to rationally design or optimize therapeutic proteins that have specific desired properties. The interactions between therapeutic proteins and drug targets are visually inspected and investigated by computational methods. This approach allows us to narrow down the list of individual amino acids that contribute most to the interaction between proteins and their binding partners, allowing us to rationally improve the properties of therapeutic proteins which can then be experimentally expressed, purified and tested.