There are now over 600,000 small molecule crystal structures and 85000 protein structures available in the CSD and PDB databases and each database can be queried using an array of powerful tools. The structural information that can be obtained from these databases can answer many questions the modeler has. However structural data can be misused too.This talk will look at new ways that structural data has been applied to validation of both ligand and protein models. Some of the pitfalls will also be described and in particular we will give a comprehensive reanalysis of calculation of strain in a protein-bound ligand, and show that extreme care is needed in choice of the initial structural model if over-high estimates are to be avoided.

Use and abuse of structural data in model validation and analysis

245th National Meeting (2013)

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