Comparative analysis of molecular 3D metrics in fragment space and beyond

Abstract

Within the fragment-based drug discovery (FBDD) community, there is growing interest in developing non-planar, 3D screening fragments to complement the predominantly 2D structures that currently populate fragment libraries. Molecular three-dimensionality is typically evaluated using metrics such as the fraction of sp³-hybridized carbon atoms (F_Csp³), plane of best fit (PBF) and/or principal moment of inertia (PMI). In this study, we review these 3D metrics and perform systematic cheminformatic analyses. Furthermore, we explore the use of alternative metrics that consider atom hybridization for the assessment of molecular shape. Although our analysis is grounded in a fragment-based context, the insights gained are broadly applicable across the drug discovery phases.