For efficient structure-based drug design it is crucial to quickly assess the relevance of molecular conformations and interactions generated in a computer model. A reality check can be performed by comparing the outcome of a simulation with the wealth of experimental information that is contained in the Cambridge Structural Database (CSD). We will show how data mining of the CSD can be used to generate practical guidelines about preferred small molecule conformations, intramolecular hydrogen bonding motifs, and intermolecular interaction preferences. Examples from small molecule drug discovery projects at Roche where this knowledge was purposefully used in the design process will be highlighted.<br/><br/>Another important use of CSD data is in virtual compound or fragment libraries of drug design software tools. Prominent examples are ReCore for inhibitor scaffold replacement, TorsionAnalyzer for assessing small molecule ligand strain, and CSD-CrossMiner for pharmacophore searching in protein binding sites. We will present case studies where these tools where used prospectively to improve binding affinity and selectivity, and to generate alternative inhibitor scaffolds.

Improved structure-based drug design with one million small molecule crystal structures

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.