In silico docking of probable substrates can guide in vitro testing of enzyme specificities.  To use large-scale docking to annotate protein structures whose functions are unknown, it is desirable to establish to what extent docking can be used reliably.  In this study, we examine the use of docking as a component of an overall function prediction strategy.  In silico docking of 50 sugar acids was performed on 28 sugar acid dehydratases, including 8 dehydratases with known substrates and 20 without.  The sugar acids were then ranked according to binding energies.  For three dehydratases with known substrates, the correct substrate was among the top five ranked sugar acids, indicating that docking  alone is not a consistent method for function assignment.  THEMATICS-SALSA analysis, followed by docking, can lead to improved predictions.  We also present predictions on 20 uncharacterized, putative sugar acid dehydratases from Structural Genomics in the enolase superfamily.

Computational approach to the prediction of enzyme specificities

245th National Meeting (2013)

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