作者: Renato Buzio , Luca Repetto , Francesca Giacopelli , Roberto Ravazzolo , Ugo Valbusa
DOI: 10.1093/NAR/GKS210
关键词: Chain (algebraic topology) 、 Orientation (computer vision) 、 Curvature 、 Nanostructure 、 Characterization (materials science) 、 Biology 、 Computation 、 Product (mathematics) 、 Bent molecular geometry 、 Biological system
摘要: We propose a method for the characterization of local intrinsic curvature adsorbed DNA molecules. It relies on novel statistical chain descriptor, namely ensemble averaged product curvatures two nanosized segments, symmetrically placed contour atomic force microscopy imaged chains. demonstrate by theoretical arguments and experimental investigation representative samples that fine mapping average along molecular backbone generates characteristic pattern variation effectively highlights all pairs tracts with large curvature. The centrosymmetric character descriptor enables targetting strands unknown orientation. This overcomes remarkable limitation current strategies estimate maps solely from trajectories end-labeled molecules or palindromes. As consequence our approach paves way reliable, unbiased, label-free comparative analysis bent duplexes, aimed to detect conformational changes physical biological relevance in sample numbers. Notably, such an assay is virtually inaccessible automated computation algorithms proposed so far. foresee several challenging applications, including validation adsorption bending models experiments discrimination specimens genetic screening purposes.