The use of gene-focussed libraries for screening against protein targets can improve timelines for drug discovery projects. This is especially true when the library is based on a novel core scaffold, avoiding the potential need to scaffold hop from early hits. Identification of an appropriate novel scaffold is therefore integral to the success of such a library. In this article we outline a new method to aid scaffold design that combines structure- based virtual screening (VS) with a second phase in which fragmentation of the output is made before the final scaffold design step. Through consideration of a refined set of bound fragments, in the context of the compounds from which they originated, appropriate vectors for appended R-groups can be assigned before validation of the final library.