Evaluation of Autodock Vina for use in fragment-based drug discovery

Abstract

The most common approaches to treatment of cancer are chemo- and radiation-therapies that cause DNA damage. The efficiency of these treatments is reduced over time because cells up-regulate DNA damage response and repair. Interest has grown in developing adjuvant therapies that will suppress cellular responses to DNA damage. Replication Protein A (RPA) has been identified as a potential target for suppression of the DNA damage response. The RPA70N domain is critical to initiation of the response to DNA damage and is therefore being pursued in our laboratory. We use a fragment-based drug discovery (FBDD) approach that involves: screening libraries of small molecular fragments against a specific target, optimization of these weakly binding fragments, and then linking them together to produce high affinity molecules. My research tested the ability of the Autodock Vina molecular docking program to expedite this process by enabling in silico evaluation of ligand binding to RPA70N. The analysis involved fragments from the Vanderbilt Structure-Activity Relationship by Nuclear Magnetic Resonance library as well as some linked compounds. My results showed that Autodock Vina is able to accurately predict ligand binding sites when the binding surface on the protein is known, but not when the entire surface has to be considered. The program also was more accurate for larger versus smaller molecules. Overall, Autodock Vina appears to not be suited for general screening of fragments but could be very powerful for the design of lead-compound analogs, thereby expediting the overall speed of FBDD.