Guiding Compound Design in 2,3...N Dimensions
Streamlining Drug Discovery and Development: Leveraging Data Analysis and Modelling for Design
Dr. Matt Segall, CEO, Optibrium
Analyses of 2-dimensional (2D) structure activity relationships (SAR), such as activity cliff detection or matched molecular pairs and series, reveal important patterns in data that can guide the design of compounds with improved potency. Furthermore, 2D quantitative structure-activity relationship (QSAR) models can be used to predict many absorption, distribution, metabolism and excretion (ADME) and physicochemical properties that are also required in a high-quality candidate drug. Where structural information is available for the therapeutic or potential off-targets, 3-dimensional (3D) methods provide understanding of ligand-protein interactions that drive activity and selectivity. However, there is often a separation between these 2D and 3D views of the world.
In this presentation we will explore how 2D and 3D SAR can be seamlessly linked, to gain synergy from the information they provide. We will illustrate how activity cliffs identified from 2D SAR analysis of experimental data, can be rationalized using 3D models to inform optimization strategies. Furthermore, docking scores or affinity estimates for virtual compounds can be used in combination with property predictions from QSAR models to achieve true multi-parameter optimization in ‘N-dimensional’ property space, while taking into account the statistical uncertainties inherent in the results of these calculations. This helps to quickly target high quality, potent compounds while ensuring that potentially valuable opportunities are not missed.