Use of Amino Acid-Nucleotide Base Pair Potentials in Screening Protein-DNA Docked Complexes

Abstract

Abstract-Amino acid-nucleotide base pair potentials are used to screen docked complexes generated by DOT. The pair potential algorithm designed in this paper is applied to screening 10 systems selected from protein-DNA benchmark set. For all the systems, a correct docking was placed within the top 6% of the pair potential score ranked complexes. Also, over 60% correct answers rank in the top 10% of the docked results for most of the systems.Index Terms-Amino acid-nucleotide base pair potentials, protein-DNA dock, screening. I. INTRODUCTIONProtein-DNA interactions regulate many cellular processes involving gene expression, DNA replication and repair [1]. Since DNA play very important roles in cells, they are molecular targets of many clinically used drugs, such as anticancer drugs and antibiotics [2]. Study on the protein-DNA interactions would be meaningful for drugs design on the nucleic acids. However, the determination of the protein-DNA structure is still difficult through the biochemistry experiment directly. Until January 5, 2013, there are over 87,000 structures deposited in the PDB (Protein Data Bank) [3], where the number of protein-NA structure is just almost 4,000. Hence, computational techniques such as protein-DNA docking will become an increasingly important way to help understand the molecular mechanisms of biological systems [4].Although a number of approaches are available for predicting the docking of protein-protein, the treatment of docking of small ligands and proteins with other proteins, the treatment of the docking of nucleic acids onto proteins lags far behind [5]. Two particular problems have hampered the development of efficient docking methods: the sparsity of the information to define the DNA-binding interface and the inherent flexibility of DNA [6]. Nonetheless, study groups such as Takeda et al. [7] and Liu et al. [8] have already made achievements on modeling protein-DNA interactions using different computational methods. However, these treatments have never applied amino acid-nucleotide base pair potentials on the docking of protein-DNA even if empirical residue-residue pair potentials [9]-[11] have already played an important role in protein-protein dockings for a long time. II. MATERIALS AND METHODS A. The Program DOTThe program DOT [12], which uses the sum of both a Poisson-Boltzmann electrostatic energy and a van der Waals energy as its energy function, quickly finds low-energy docked structure for two macromolecules by performing a systematic search over six degrees of freedom. A major objective of the DOT program is to provide a method that is fast enough for routine use and cheap enough to be used in highly speculative modes.In our earlier work, we applied DOT to dock 10 systems selected from protein-DNA benchmark set [13] and the results showed DOT was able to produce correct answers efficiently. However, these results were not ranking forward enough to guide the design of experiments to test the suggested interactions. Hence, amino acid-nucleotide base pair potentials w...