SAR matrices provide a novel data structure to automatically extract SAR patterns from data sets and organize compound sets in a chemically intuitive scaffold/functional-group SAR table view. This computational approach uses a matched molecular pair-like algorithm to identify and automatically extract groups of structurally related compounds, and displays the resultant information in a chemically interpretable fashion. SAR matrix content can be used to capture SAR discontinuity, identify areas which will require further SAR exploration, or suggest virtual compounds for synthesis. This framework also delineates areas (identifies compound subsets) in (for) which computational techniques are reasonably predictive. Various examples of SAR matrices have been generated and studied in order to interrogate the existing wealth of information contained in project data. Specifically, applications of SAR matrices will be presented in the context of 1) over-explored or under-explored scaffolds; 2) privileged R-groups; 3) activity cliffs; and 4) appropriate holes to fill in SAR matrices.

Novel applications of SAR matrices in pharmaceutical research

247th National Meeting (2014)

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