Quantifying Sources of Error in QM Computations of DNA Intercalation

DNA Intercalators have been clinically used as anti-tumor, anti-parasitic, and anti-bacterial agents. One method used to discover novel intercalators involves screening large virtual libraries of drug-like molecules with docking programs. In general, these programs dock flexible ligands into a rigid receptor and then rank binding poses based on a qualitative scoring function. However, the binding of DNA intercalators is not described well by these molecular-mechanics-like scoring functions due to the small distances between the intercalator and DNA bases. I will first describe computational studies aimed at understanding which available quantum mechanics-based methods provide the most accurate interaction energies.  Second, I will assess how much of the DNA strand needs to be retained in computational models in order to provide accurate predictions of intercalation.