Implementation And Comparison Of Model For Bio-molecular Helices And Research On Characteristic Of DNA-binding Protein

The visualization of the bio-molecular for the study of bio-molecular structure,especially some molecular whose structure information has a key role in the function of thelife, such as protein and DNA, has a very important biological significance. Some importantinformation about the structure and function of proteins is often directly inferred by observingclear and intuitive three-dimensional images of protein molecules. The structure of DNAmolecules is a typical double stranded helix, but the helix structure will be distorted whenbinding happens between other molecules such as protein and DNA molecules. The distortionoften happens in the binding sites. It is necessary to analyze the structure and function ofbiological molecules and discover the laws of the structure and function by visualizing thestructure of bio-molecular and the distortion in the binding sites with visualization programs.The helix is the common geometry feature of many bio-molecular. The α-helix structureis one of the protein structures; DNA is a typical double stranded helix. One of the functionsof the ribbon model is to show their helix structure accurately. But in the past, a bio-molecularhelix in either a protein or a DNA is generally not modeled as a genuine helical curve. Insteada bio-molecular helix is usually approximated with a series of splines such as low degreeHermite polynomials that pass through backbone C or P atoms. This method could deviatelargely from a genuine helical curve at both the local and global levels, so we need a newmodel to draw a bio-molecular helix structure.In this paper, we implement a new model that represents a bio-molecular helix with anewly-developed algorithm, compare with the ribbon diagrams drawn by other molecularvisualization programs and analyze their advantages and disadvantages. The model uses acurve fitting algorithm that searches for a helical curve that best fits the coordinates of fourbackbone atoms. The algorithm defines a new score to evaluate the helix structure andquantify the local deviations and to further link such deviations with protein-ligand bindingsites. The distortions could be easily observed through the helix diagram drawn by this model.When the model is applied to the visualization of a helix as a ribbon diagram and compared with other molecular visualization programs, it could eliminate the choppiness in proteindiagrams and to greatly reduce it in DNA diagrams. The minimization of theroot-mean-square deviation (RMSD) between four backbone atoms and their closest point onthe diagram achieved by the curve fitting algorithm greatly alleviates the side chaindetachment problem.Another research theme of this paper is the study of the characteristic of DNA-bindingprotein interface. An understanding of the mechanism of the interaction between DNA andprotein for the study of the physiological phenomenon of the living things has a veryimportant significance. In this paper we calculate and analyze the hydrophobic area of theDNA on the interface, develop the law of the statistics, and provide the basic data for thefurther study of the mechanism of the DNA-protein interaction.