Investigating Influence of Conserved Residues on Binding Specificity Based on Existing Protein-DNA Complexes

The functions of about half of the genes are still unknown for most species of which the genomes are already sequenced. Many of these genes encode transcription factors (TFs) which can recognize target genes. Some databases provide information about transcription factor binding sites from experiments. Such data are generally used for searching DNA sequences targeted by the transcription factors. However, these experiments consume much cost and human power. Obviously, it is necessary to combine as much information as much as possible to discover TF targets. Increase of data on the structures of protein-DNA complexes provide us with more information about the specific interactions between DNA bases and amino acids. It is observed that there is no simple one-to-one correspondence between bases and amino acids. The type and orientation of the interactions are extensively spread in space. More than one possible geometric relationship between bases and amino acids geometry would be generated.The objective of this thesis aims to investigate the effects of conserved residues on binding specificity from the aspect of energy. We use the program, Rosetta, to estimate the binding energy of protein-DNA complexes. By exploiting the data from Protein Data Bank (PDB), we put the coordinate information of each atom as the input to Rosetta. Rosetta can refine the 3-dimension coordinates of all the atoms from the PDB file for optimal energy performance. After calculating the energies of the complex before and after the mutations of bases or residues we define every base as causing remarkable or non-remarkable energy changes to complex. With conserved and non-conserved residues in the same complex we have two types of characters, conserved and non-conserved residues, remarkable base and non-remarkable base. Finally we employ statistical analysis to evaluate the influence of conserved residues on binding specificity.