computational structural analysis multiple proteins bound to dna计算多个蛋白质绑定到dna结构分析.pdf

Computational Structural Analysis: Multiple Proteins Bound to DNA Andrija Tomovic*, Edward J. Oakeley Friedrich Miescher Institute for Biomedical Research, Novartis Research Foundation, Basel, Switzerland Abstract Background: With increasing numbers of crystal structures of protein:DNA and protein:protein:DNA complexes publically available, it is now possible to extract sufficient structural, physical-chemical and thermodynamic parameters to make general observations and predictions about their interactions. In particular, the properties of macromolecular assemblies of multiple proteins bound to DNA have not previously been investigated in detail. Methodology/Principal Findings: We have performed computational structural analyses on macromolecular assemblies of multiple proteins bound to DNA using a variety of different computational tools: PISA; PROMOTIF; X3DNA; ReadOut; DDNA and DCOMPLEX. Additionally, we have developed and employed an algorithm for approximate collision detection and overlapping volume estimation of two macromolecules. An implementation of this algorithm is available at http:// promoterplot.fmi.ch/Collision1/. The results obtained are compared with structural, physical-chemical and thermodynamic parameters from protein:protein and single protein:DNA complexes. Many of interface properties of multiple protein:DNA complexes were found to be very similar to those observed in binary protein:DNA and protein:protein complexes. However, the conformational change of the DNA upon protein binding is significantly higher when multiple proteins bind to it than is observed when single proteins bind. The water mediated contacts are less important (found in less quantity) between the interfaces of compo