Grafting – Predicting Specificity

The Grafting Webtool




Welcome! The Grafting Webtool allows researchers to predict glycan-binding protein (GBP) specificity using computational carbohydrate grafting. It is possible to perform “virtual glycan array screening”, and to validate putative 3D structures of co-complexes based on their agreement with experimental specificity data. Please refer to our publications for examples of what grafting can do.

General overview of grafting
Given a 3D structure of a carbohydrate fragment (i.e. a minimal binding determinant) in complex with a GBP, the webtool will find full glycans that contain this fragment and build them into the binding site using the known conformational preferences of carbohydrates. If it is possible to fit the larger glycan into the binding site, it is predicted to be a binder.

The webtool can model glycans present on microarrays or found within the human glycome, allowing prediction of GBP specificity. Conversely, a user can validate a modelled 3D complex (e.g. generated via by comparing the predicted specificity to experimental binding data. Following the tutorials is a great way to get started. Before running your own system, please read the instructions!

Please send issues/suggestions/requests etc to, including input files if relevant. User feedback will be vital to make this a successful tool, and we are grateful if you take the time to write us.