Strong Nonadditivity as a Key Structure–Activity Relationship Feature: Distinguishing Structural Changes from Assay Artifacts

• DOI: 10.1021/acs.jcim.5b00018
• Publication Date: 2015-03-11T16:24:59Z
• Authors: Christian Kramer; Julian E. Fuchs; Klaus R. Liedl
• Keywords: ERG1 Potassium Channel; 0303 health sciences; Uncertainty; Proteins; Crystallography, X-Ray; Ligands; Ether-A-Go-Go Potassium Channels; Structure-Activity Relationship; 03 medical and health sciences; Factor Xa; Cytochrome P-450 CYP3A; Estrogen Receptor beta; Matrix Metalloproteinase 2

Nonadditivity in protein–ligand affinity data represents highly instructive structure–activity relationship (SAR) features that indicate structural changes and have the potential to guide rational drug design. At same time, nonadditivity is a challenge for both basic SAR analysis as well many ligand-based techniques such Free-Wilson Analysis Matched Molecular Pair analysis, since linear substituent contribution models inherently assume additivity thus do not work cases. While causes been analyzed anecdotally, no systematic approaches interpret use prospectively developed yet. In this contribution, we lay statistical framework of series. First, develop general metric quantify nonadditivity. Then, demonstrate non-negligible impact experimental uncertainty creates apparent nonadditivity, introduce handle uncertainty. Finally, analyze public sets strong recourse original publications available X-ray structures find explanations observed. We all cases (ΔΔpKi ΔΔpIC50 > 2.0 log units) with sufficient information generate reasonable hypothesis involve binding mode. With appropriate basis, offers variety new attempts various areas computer-aided design, including validation scoring functions free energy perturbation approaches, pocket classification, novel tools.