Insights into RNA recognition of small molecules using the Fragment Molecular Orbital (FMO) method

Predicting the binding affinity between a ligand and a target is the key to optimizing target-ligand interactions and designing bioactive drug-like molecules. Unfortunately, current state-of-the-art empirical methods, when applied to RNA-ligand complexes, exhibit weak correlations between predicted and experimentally measured binding affinities between the RNAs and ligands. Fragment Molecular Orbital (FMO) method1,2 is a divide-and-conquer technique that enables ligand binding energies to be computed using a high-level quantum mechanical (QM) description of the interactions between a ligand and a target receptor.3 Here we describe, to the best of our knowledge, the first reported use of FMO calculations to predict binding energies in RNA-ligand complexes4. We found that FMO calculations could be used to generate reliable estimates of the relative interaction strengths in RNA-ligand complexes. Also, using these FMO calculations, we were able to decompose the binding energies into its specific components and identify component(s) that are dominant. If our preliminary results hold, FMO calculations should find utility as a high-accuracy yet general approach for studying the molecular recognition of small molecule ligands by RNAs associated with human diseases, and in so doing aid is structure-guided design of RNA chemical probes.